Thursday, March 22, 2018

Update Source Anchor to ms-DS-ConsistencyGUID

A key part of most Office 365 deployments is directory synchronization withe on-premises Active Directory (AD). To maintain a link between individual object in AD and Office 365, one attribute in AD is defined as the source anchor. The source anchor acts as a unique identifier for each object so that you can change properties like UPN and have them replicate to the proper matching object in Office 365.

Older versions of Azure AD Connect used the objectGUID attribute as the source anchor. In many cases, this is a good choice because it is an attribute that doesn't change. However, in complex configurations with multiple AD forests, this can cause an issue during migrations. When a user account is moved from one AD forest to another, it gets a new objectGUID in the new forest. Due to the new objectGUID , Azure AD Connect treats the user as a new user instead of an existing user.

To avoid this issue, we need to use an attribute for source anchor that can be moved between forests. Microsoft now recommends that deployments of Azure AD Connect use the mS-DS-ConsistencyGuid attribute as the source anchor. Unlike objectGUID , this attribute can be modified. So, it can be maintained when users are moved between forests.

I've just recently completed this process and I'll provide some general guidance (for my own reference) and a few items that I didn't see anywhere else.

Updating the Source Anchor

First, read the information from Microsoft on updating the source anchor. You'll need to scroll down a bit to the heading "How to enable the ConsistencyGuid feature - Existing deployment".
This documentation provides click by click instructions on how to use the Azure AD Connect configuration wizard and work through the Change Source Anchor option. This wizard will:
  • Require you to enter Azure AD credentials for a global admin
  • Require you to enter a user account with AD FS admin permissions (if integrated with AD FS)
  • Do a prerequisite check
  • Update necessary settings in Azure AD
  • Update necessary settings on connectors for all connected forests
  • Update AD FS claim rules
If you're like me, before attempting this you'll scour the web looking for information about how to fix things when they go wrong. Unless things change a lot after I write this, you won't find much. The reason you won't find much is because the wizard does exactly what it's supposed to.

Note: The wizard does not ask you which attribute you want for the source anchor. It is assumed that you want to use mS-DS-ConsistencyGuid.

Change Details

When the wizard is done, a full sync is triggered for all connected forests. As part of the full sync Azure AD Connect reviews the value of the mS-DS-ConsistencyGuid attribute. If the value is blank, the objectGUID value is copied into the mS-DS-ConsistencyGuid value. Then on the next sync cycle the mS-DS-ConsistencyGuid value is sent to Office 365.

If you are using AD FS, the wizard updates the AD FS claims rules for Office 365. I didn't do a detailed review of the rules, but it seems to be similar (if not the same) as the rules Microsoft documents for you to make manual changes. You can see these rules here:
The Modify AD FS Claim Rules section in the above link has four rules. The net effect of these rules is to use mS-DS-ConsistencyGuid if it is populated and use objectGUID otherwise. This avoids downtime when AD FS rules have been modified but the mS-DS-ConsistencyGuid hasn't been populated yet.

Permissions to mS-DS-ConsistencyGuid

If you have let the installation and configuration wizards for Azure AD Connect manage all of the service account permissions and AD FS integration, there should be no issues. However, if you've been manually configuring the permissions and AD FS rules, you might need to make changes manually.

The only permission change required when changing the source anchor is ensuring that the service account in each domain has read and write permissions to the mS-DS-ConsistencyGuid attribute. Unfortunately when you attempt to assign these permissions by using Active Directory Users and Computers, you find out that the attribute is not displayed in the GUI for modifying permissions. To set the permissions, use the following command (please note that this is a single line):
dsacls.exe "dc=contoso,dc=com" /I:S /G "contoso\serviceaccount:RPWP;mS-DS-ConsistencyGuid;user"
The above command grants (/G) inheritable (/I:S) permissions to the service account to read (RP) and write (WP)  the mS-DS-ConsistencyGuid attribute on user objects.

Verifying Successful Implementation

Obviously after a successful implementation, your users can still sign in to Office 365, but there are a few other things you can verify:
  • Review the AD FS claim rules for Office 365 and verify that they were modified.
  • Review the mS-DS-ConsistencyGuid attribute on several user objects to verify that it has been populated.
On one deployment, we had an error in the Azure AD Connect full sync for one domain immediately after finishing the wizard. The domain that finished properly updated the mS-DS-ConsistencyGuid attribute and the domain with the error didn't. Running an incremental sync didn't fix it. We needed to force a full sync again to get the attribute updated properly.

When viewing the value of the mS-DS-ConsistencyGuid attribute in Active Directory Users and Computers, the value might be displayed in special characters for some users. This is a display error in Active Directory Users and Computers. This is not something you need to fix.

For Further Reading

Tuesday, March 13, 2018

O365 Removing Mobile Device Fails in Portal

A client has an issue in their Office 365 tenant that started yesterday (March 12/18). When attempting to delete a mobile device linked to a user, it fails and gives the following message:
Error on proxy command "Remove-MobileDevice -Identity:'DeviceName' -Confirm:$false to server...the mobile device DeviceName cannot be found...

Error when removing mobile device
I was still able to remove the device by using a Windows PowerShell prompt connected to Exchange Online. But what I noticed is that for the Remove-MobileDevice cmdlet to work, I needed to use the Identity property from the mobile device rather than the Name property that the Exchange admin portal was attempting to use.

Using Windows PowerShell to remove the mobile device
You can see in the screenshot above using the Name or DeviceID properties didn't work. Only the Identity property which includes the full path with the user name works.

I've opened a support ticket with Microsoft and we'll see what they come back with. It was not a know issue when I contacted them, but they're going to do some checking on the back end. I will update this blog post when there is a resolution.

UPDATE: March 14/18
So, support indicates that it's definitely an issue, but to raise visibility with the support team they asked me to create an "idea" at This location seems more for feature requests than bug fixes. However, I've put an "idea" in. I'll continue to update as I learn more.

If you are also experiencing this issue and want to raise its visibility, go to this link and vote it up:

Saturday, January 27, 2018

Outlook on the web offline access failure

There is a bug in Exchange Server 2016 CU6, CU7, and CU8 that prevents clients from using offline access for Outlook on the web. It was working properly in CU5.

In CU5, in Options, you select Turn on offline access and then another window pane is shown that asks whether you have exclusive use of the computer or not. In CU8 (also CU6 and CU7), that pane appears very quickly and then goes away. Basically the text flickers but it is removed right away. You can see a video of this behavior below.

I tested this with Internet Explorer, Edge, Firefox, and Chrome with consistent results. I also tested this using both Windows 10 and Windows Server 2016 as the client operating systems.

After trying everything I could think of for testing, I posted this issue to the Microsoft support forums and they confirmed that this is a bug and will be fixed in a future update.

Wednesday, January 24, 2018

Install-AntispamAgents.ps1 Fails for Exchange 2016 CU7 and CU8

Not many organizations enable the antispam agents on their Exchange servers since they're probably paying for an additional service to do spam filtering. However, if you attempt to enable it on Exchange Server 2016 CU7 or CU8, there is an error in the script that you need to run. As of right now, CU8 is the latest update available so, this may continue in later updates.

To enable the antispam agents in Exchange Server 2016, you run the Install-AntispamAgents.ps1 script located in C:\Program Files\Microsoft\Exchange Server\V15\Scripts. However in CU7 and CU8, you get an error like this:
There are several screens of error information, but the key part is:
A parameter cannot be found that matches parameter name 'EscalationTeam'.
If you open the script and go to line 50, you'll see the following at the end of the line:
If you delete the following text from that line, the script works properly:
-EscalationTeam "AntiSpam";
Please note that this script does have a digital signature. Editing the script invalidates the digital signature. So, to run this script ensure that your execution policy is not set to AllSigned. It will run fine with execution policy set to either Unrestricted or RemoteSigned.

Monday, January 22, 2018

All Certificate Names MUST be in Subject Alternative Names

This has been popping up for a while, but it's worth pointing out again. When you get a SAN/UCC certificate, the DNS name that you use for the subject (common name) also needs to be in the subject alternative names attribute.

For example:
  • Subject:
  • Subject alternative names
When you get your certificate from a public certificate authority (CA), most of them ensure that the subject is also added as a subject alternative name. However, some might not and you should watch for it.

It's important to note that the Microsoft CA does not automatically add the subject to the list of subject alternative names. So, make sure that you do it as part of your certificate request for your internal CA.

The reason you need to do this is because of how web browsers are processing subject alternative names. Web browsers are ignoring the subject if a list of subject alternative names is present. All web browsers seem to be enforcing this now. A few years ago, the web browsers would process both the subject and the subject alternative names.

Here's the quote from RFC 2818 from May 2000 (yes, that long ago):
If a subjectAltName extension of type dNSName is present, that MUST be used as the identity. Otherwise, the (most specific) Common Name field in the Subject field of the certificate MUST be used. Although the use of the Common Name is existing practice, it is deprecated and Certification Authorities are encouraged to use the dNSName instead.

As a slightly interesting side note. The subject field is optional when SANs are defined. However, some older software might still require the subject to be defined.

If you are configuring certificates in Exchange server, be aware that the Exchange admin console (EAC) does not properly differentiate between subject and subject alternative names. In the certificate shown below, is in the subject and not in subject alternative names. EAC (Exchange 2016 CU8) is including the subject in the Subject Alternative Names box which can be misleading.


Tuesday, December 19, 2017

PowerShell cmdlet to query .NET Framework version

Whenever I'm doing new installs of Exchange, I'm annoyed when I need to figure out the .NET Framework version installed based on registry keys. You can see the documentation here:
 Just like the bad infomercial says: "There's got to be a better way!"

This has bothered me for sometime and I considered writing a small script to query and display the version. Then I thought...that would be good if I put it in the PowerShell Gallery so that I could easily access it from anywhere. Then I thought...I'm probably not the first person to have this idea.

In the PowerShell gallery, Joakim Borger Svendsen has been kind enough to create this for us as a module named DotNetVersionLister. He's also kind enough to keep it updated. There have been seven releases in 2017.

On a computer that has Internet access, you can install directly from the PowerShell Gallery:
Install-Module DotNetVersionLister
You can also save it to file for distribution to computers that don't have Internet access by using:
Save-Module DotNetVersionLister -Path C:\Temp
After you install the module, you have a single cmdlet named Get-DotNetVersion. This cmdlet is quite flexible and allows you to query the .NET Framework version from the local computer and remote systems. For remote systems, it can query the information using remote registry access or PowerShell remoting. So, it's flexible depending on your configuration.

Query the .NET Framework version from the local computer:
Get-DotNetVersion -LocalHost
Query the .NET Framework version from a remote computer via remote registry:
Get-DotNetVersion -ComputerName Exch01
Query the .NET Framework version from a remote computer via PowerShell remoting:
Get-DotNetVersion -ComputerName Exch01 -PSRemoting

Related links:

Tuesday, December 12, 2017

Docker Fails to Start on Reboot

For the last couple of weeks I've been working with Windows containers using Docker. I ran into a severe problem with the networking. I created a transparent network on the host and rebooted. After reboot, the docker service wouldn't start and had the following error in the event logs:
Log Name:      Application
Source:        docker
Event ID:      4
Task Category: None
Level:         Error
Error starting daemon: Error initializing network controller: error obtaining controller instance: failed to get endpoints from store: failed to decode endpoint IPv4 address () after json unmarshal: invalid CIDR address:
My first fix for this issue was to delete the following file:
  • C:\ProgramData\docker\network\files\local-kv.db
After this file was deleted, I was able to start the Docker service and it stayed running. That file was recreated when the docker service started and I was able to run docker commands.
Running docker network ls showed me that a transparent network I had created just before the restart was broken. That network was renamed a long random string. At this point, I could delete the randomly named transparent network, but a new one came back after each restart of either the Docker service or the host.

The final fix to stop that recurrence was running:

.\WindowsContainerNetworking-LoggingAndCleanupAide.ps1 -Cleanup -ForceDeleteAllSwitches

That script is provided by Microsoft on GitHub here:
It's also worth noting that others are also having this issue: