ADFS WAP behind Azure Application Gateway

Microsoft Azure

Some time ago i wrote up a post (located here) explaining how you can setup traffic manager with ADFS and have proper monitoring of the service. Today i will go over how to setup ADFS behind the Azure Application Gateway. This will enable you to protect your ADFS service and monitor it with the WAF provided by the application gateway.

Before we begin one prerequisite which i am still not sure if its really needed but i had problems and i believe this fixed it:

You need to set the default HTTPS Binding, i believe this is required as i am not sure if the health probe is truly SNI compliant, i might be wrong here but it doesn’t hurt to set this. To set it you simply need to run the following command on the WAP servers (just change the cert hash):

Ones that’s done create a Application gateway in Azure and do the following:

  1. Create a Frontend listener with thew following settings:
    • HTTPS Protocol
    • Listen on port 443
    • Multi-Site type, you can do basic but that will limit your application gateway to only the ADFS service for port 443
    • Provide a PFX file of your ADFS certificate. make sure you include the private key and a strong password
  2. Create a Health Probe with thew following settings (just change the host):
    • The path (so you can copy and paste): /adfs/ls/IdpInitiatedSignOn.aspx
  3. Create a HTTP Setting with thew following settings
    • HTTPS Protocol
    • Cookie based affinity: Disabled (you really don’t need that for ADFS)
    • Port 443
    • Export your ADFS certificate as a base 64 format (do not include the private key) and add it.
    • Tick the “Custom probe” and select the probe we created earlier
  4. Create a Backendpool which includes all your WAP servers
  5. Crete a Basic Rule using the objects created earlier.

And that’s it, this is not only a secure solution but it will give you a proper monitoring of both the WAP and ADFS servers. Works great with loadbalancing between on-prem and Azure.

Azure AD Authentication (Connect-AzureAD) in Azure Automation

Microsoft Azure

It is now (has been for a while) possible to modify Azure AD via the Azure Automation. The example below uses the Run As Automation Account to first Connect to Azure AD and then run the appropriate commands. You can also create a dedicated Run As account if you want, as well as use a username and password (less secure).

Before you write your code make sure that you:

  • Add the “AzureAD” module to the Automation Account
  • Give the Azure Automation Run As account the appropriate permission as show at the end of this article

Automation Code example (list all the groups in AD):

Give the Azure Automation Run As account the appropriate permissions:

  • Go to Azure Active Directory -> App registrations -> The Run Ass Account.
  • Then go to the API access as show:

  • Give the appropriate access, example below:

Don’t forget to click grant permissions!

Azure ASR Error- 78052 Master target contains different types of scsi controllers.

Microsoft Azure

This is a bit of a self-explanatory one, but I thought I would mention it anyway. When you build an ASR Master Target server make sure if you have more than one SCSI controller that they are of the same type, it doesn’t matter what type they are (LSI Logic SAS, VMware Paravirtual, ect..) but they both need to be the same or you will get the following error on the Azure portal when you attempt to fall back the machine to On-premeses.


Azure ASR Error- 90068 disks specified not present

Microsoft Azure

Quick fyi for anyone using Azure ASR, make sure if you are protecting a virtual machine located in Azure to unselect the temp drive disk D when you are adding the machine to ASR protection. If you try and protect the disk to on-premises, you will get the below error message. If you do you will need to delete the protection and reprotect without drive D. The below error only occurs when you try and reprotect to on-premises, it seem to work fine if you reprotect to another azure location.


Traffic Manager Endpoint monitor and ADFS /adfs/probe

Microsoft Azure, Windows

Microsoft has a very nice post on how to setup Traffic manager in front of an ADFS farm for high availability, where both sites are in Azure but in different GEO locations or one in Azure and one on premises. The Article is located here: What the article lacks is how to setup proper ADFS monitoring, which monitors both tte WAP and the ADFS service, at the moment the article only goes into details which monitor the WAP service.

So this post will go over how to configure your environment so the health point will report the status of both WAP and ADFS.

Some info before we begin:

  • The solutions is achieved by monitoring the /adfs/probe/ on the ADFS server via the WAP proxy
  • The solution will report failure if the WAP proxy is not forwarding or the ADFS service is down. So we are monitoring the whole solution.
  •  It will work if you have an external load balancer in front of the WAP servers and an internal one in front of the ADFS servers, for simplicity I will outline how it’s done on the non-load-balanced solution but it’s the same procedure for both.
  • You can’t monitor /adfs/probe on the WAP server as that will only give you the status of the WAP server
  • You can create a rule on the WAP server to redirect /adfs/probe to the ADFS server, but it will get ignored and show you the status of the WAP server.
  • I tested this on Server 2016 but it will work for 2012 R2 as well
  • If you are using 2012 R2 make sure you update your WAP to the latest version so you can forward HTTP traffic
  • We use HTTP as this prevents certificate problems and because Traffic manager does not support SNI.
  • You can’t monitor the “/federationmetadata/2007-06/federationmetadata.xml” because the way you set this up for Traffic manager means you are monitoring the ADFS on a different DNS so the request will not be forwarded.

Essentially this is what we are doing


Once you setup the environment as per Microsofts Article above we need to do the following:

The variables for my test environment:

  • ADFS URL and Federation Service Name – test123.blah.local
  • Traffic Manager DNS –
  • WAP server public IP dns (this can be replaced by a load balancer) –
  • Custom monitor path (you can choose anything but the default which is /adfs/) –  /adfsprobe/

The Steps:

  • Change the Traffic Manager Configuration to point to our custom monitor path for the endpoint monitoring


  • Create an HTTP rule on the WAP server in the Remote Access Management Console to forward (via Pass- through) the WAP DNS + our custom monitor path to the ADFS server. I assume that your WAP server host file has been modified to point the ADFS URL to the ADFS internal IP or load balancer IP



  • The rule to be created is Reverse Proxy with the following settings:


  • And finally change your Public DNS record and create a CName for your ADFS URL (test123.blah.local) to point to the traffic manager DNS name (

And you are done.

Powershell Add-Computer error when executed remotely.


When you execute the PowerShell command: “Add-Computer -DomainName “” -Credential $domainjoinuser -Restart” remotely or in a non-interactive environment you may get the following error:

The root of the problem is (given that your password is correct) when running things interactively the domain is pre-appended and as such you only need to provide the user. But in a non-interactive environment, the domain is not known as such it’s a very simple fix, make sure you either include the short domain names like “contoso\DMAdmin” or the full FQDN “ The error occurred for me by running an Azure custom script which called a PowerShell script non-interactively.

The ACL RemoveAccessRule Not Working


If you try and modify the ACL via PowerShell but the command RemoveAccessRule is not working, by that I mean you run it no errors come up but the rules and not being removed.  The problem is that inheritance is turned on and you are trying to remove a rule that is obtained from inheritance. To fix this problem you first need to disable inheritance, save the rule and then get the acls again. After that remove will work. Code can be found below:


New Virtual Disk Error “UseMaximumSize” (Storage Pool)


This is a heads up post for anyone who is creating a new virtual disk from a storage pool in Storage Spaces. If you are creating a disk and want to use the “Maximum size” parameter then you have to make sure that the Provisioning type is set to Fixed and not Thin. This is a normal expected behaviour as the Thin disk expands automatically and take space as it needs so you can’t pre-located 100% of the size, you will need to specify an initial disk size. For fixed size “-ProvisioningType Fixed” you can use the “-UseMaximumSize” parameter.

The unfriendly Powershell Error:

The GUI Error:


Catching Azure Automation Child runbook Errors

Microsoft Azure

Today I had a very fun time trying to figure out where errors were coming from my child runbook when it ran fine on it’s own. Its seems Microsoft error handling from inline calls to child runbooks is a bit buggy and cost me a day of looking through working code to figure out the problem. It reports weird errors out of which not all are actually true or correct. You may get errors like:

If you do look at the code the foreach/if loop error is referring to as the problem is most likely coming from a function or a method inside the loop. I found deleting (not commenting as it doesn’t work) all code below the if statement or foreach loop gives a better and more precise message exactly where the error is coming from.

New-Object PSObject Problem with Azure Automation

Microsoft Azure

I found a weird behaviour in Azure automation, if you use “New-Object PSObject” it runs perfectly fine when the script is executed directly (via the portal or the Powershell command “Start-AzureRmAutomationRunbook”). However if you try and do an inline call from another workbook it fails and errors out with something like:

the code i used initially was (snippet):

It seems like this command is no longer supported or runs correctly when the child workbook is called by another workbook. To get around this i recommend using the Powershell v3 method which works perfectly: