Finding All Tenants Unassigned Office 365 Licenses

I previously wrote about finding all the mailboxes across your Office 365 tenants, and today I am going to share a function I wrote to find all the unassigned Office 365 licenses across those tenants.  First, I should note that this function is modified off of one I found on the TechNet script gallery but for some reason even after searching through it I cannot seem to find it.  I am also using the MSOnline module for this, even though the new AzureAD module has supposedly replaced it. I have been in contact with someone on the team at Microsoft about what the current limitations are.  However, the functionality doesn’t exist in that module (yet) to get the information that we need so I am continuing to use the MSOnline Module instead.

It’s a pretty straight forward function so I am just going to break down what it does:

  • Connects to Office 365.  You will need to use an account that has delegated admin permissions across your tenants
    • If you have MFA enabled you won’t be able to login (or at least I couldn’t get it to work with my account)
  • Gets a list of all the tenants you have delegated admin access for
  • Creates an Array to store all the information
  • Gets all the associated AccountSkus from each tenant, with the exception of Power_BI and PowerApps
  • Compares the value of ConsumedUnits to ActiveUnits.  If the amount of ConsumedUnits is less than the amount of ActiveUnits, we know that there are unassigned licenses.

That’s all there is to it.  Full function is below.


What I Learned About VSCode At The PowerShell Summitt

I am going to assume that if you are reading this article that you know what Visual Studio code is and have installed the PowerShell extension for it, so this article isn’t going to tell you how to do any of that. If you don’t know what VS Code is, it is a free, open-source code editor developed by Microsoft for Windows, Linux, and macOS that is completely customizable. You can read more about it on the official site and on the “Why Visual Studio Code?” page. I thought Visual Studio code was awesome before the I attended the PowerShell Summit, and I think it’s even more awesome now.

One last note before I get started. If you are reading this article it’s probably safe for me to assume that you use PowerShell and/or Visual Studio Code. If either of those things are true, you should seriously look into attending the PowerShell + DevOps Summit in 2018. This year was the 3rd year I have attended and every year it gets better. The summit is intentionally kept small, so tickets go fast. To get notified about the 2018 Summit you can sign up for a mailing list or you can follow the official Twitter account @PSHSummit.

Alright, let’s get down to business.

I knew that VS Code added PowerShell language support a while ago, what I didn’t know is that you could set the default language for VSCode to PowerShell. Here’s how you do it:

  1. Use Ctrl + Shift + P to bring up the Command Palette (you should get used to doing this).
  2. Type User Settings and click on Preferences: Open User Settings
  3. In the box on the right hand side type “files.defaultlanguage”: “powershell”
  4. Save it. Congratulations your default language for all new files is now PowerShell!



The next thing I learned is that VS Code can handle automatic code formatting for PowerShell.  Don’t believe me?  Here is some crappy unformatted code:

To have VS Code do automatic formatting you can either right click anywhere in the document and then select Format Document or you can highlight the code you want to format, right click, and select Format Selection.  Either way you end up with formatted text.

You can also completely customize your own code formatting if you don’t like the default settings.  Go into your User Settings again (just like you did to set your default language) and type “codeformat” by the time you type all of that you should see it display the available code formatting options:

Select any of those and then set the appropriate option.

Since there are squiggly lines in the code example above, now is a good time to mention the next thing that I learned.  When writing PowerShell in VS Code, behind the scenes it is using PSScriptAnalyzer to analyze the code and make recommendations.  There are the different ways to see the problems and recommendations in VS Code.  The first way is if your cursor is on a line with a squiggly, hover over the squiggly line to see the recommendation.

Also, to the left of the code you will see a little light bulb icon.  Left clicking that icon will also give you some information (similar to what hovering over the squiggly line will get you).

Finally, you can view all the “problems” that PSScriptAnalyzer has detected in one place in the output pane.  At the bottom of your screen you should have an area that looks similar to this.  If you don’t, you can enable it by going to View -> Output, or Ctrl+Shift+U.

If you click on the Problems tab, you will see all your “problems” in one place.

Next up, is creating a Visual Studio Code profile, which works exactly like a PowerShell profile.  You will seen in the screenshot showing the output window that my prompt is a $ (because PowerShell is money friend!).  How did I do that?

  1. Browse to your user documents folder and then to the WindowsPowerShell folder (C:\users\jacob\documents\WindowsPowerShell)
  2. Create a file called Microsoft.VSCode_profile.ps1
    1. Clearly you can do this through PowerShell
    2. Put whatever you want VS Code to load into your profile
  3. Reload VS Code by typing Ctrl+Shift+P and then Reload
  4. If you type $profile in the VS Code terminal window you will see an output similar to this

Next up is themes  This seems pretty obvious in retrospect but there are all kinds of themes out there you can use to customize your editor experience.  To browse through the available themes in VS Code click the Extensions icon in the left hand toolbar (square icon at the bottom) and in the search extensions in marketplace box just type Themes.  This will give you a giant list of themes but no sorting.  What about most installed themes?  Type themes @sort:installs .  What about most popular?  Type themes @sort:rating-desc .  The screenshots in this blog are from the built-in Dark theme (highlighted in the screenshot below).  You can view all the available color themes by opening up the command palette (Crtl+Shift+P), type Themes and then select Preferences: Color Theme and then browse through the available options using the up and down arrows.  Once you install a theme extension and reload VS Code it will populate into this list.

Extensions work in exactly the same way.  In the search box on the Extensions section you can type @sort:installs to see the most installed extensions (PowerShell is currently #9).  The extensions I have installed are shown below.  Azure Resource Manager Tools gets you intellisense and snippets for writing ARM Templates in VS Code, Settings Sync allows you to sync your VS Code settings across devices using GitHub, Visual Studio Code Git Hub Support allows you to save and commit code changes to GitHub (which is an entire blog post on its own) and vscode-icons gets you cool icons in the explorer window.

Finally, here are some other user settings you can set that I also learned about at the PowerShell Summit:

  1. editor.fontfamily
  2. editor.fontsize
  3. extensions.autoupdate
  4. editor.formatonsave
  5. editor.formatonpaste
  6. terminal.integrated.fontsize
  7. terminal.integrated.fontfamily

You can read about some more tips and tricks from the VS Code GitHub page.

What are some of your favorite VS Code tips and tricks?


Finding All Client Mailboxes in the Office 365 Partner Portal

Someone at work asked me “how many client mailboxes are we supporting in Office 365?”. It’s easy to tell how many clients you are supporting, as soon as you login to the partner portal it tells you that number. I had never tried to get the number of client mailboxes before, but I strongly suspected that I could do it using PowerShell. Turns out I was right!

The first thing you will need to do is to connect to Office 365 using PowerShell. If you haven’t done this before, go to this page and download the Office 365 PowerShell Module and install it.

After that is done, execute the commands below which will connect you to Office 365. It’s important to note that the credential you use must have delegated admin access to all the client accounts or the steps in this article won’t work.

Once that is done running the commands below will provide a list of all the commands in the MSOnline Module as well as only those commands which contain the word partner. I was hoping there was a cmdlet that would allow me to get all my clients information which I could then iterate through to get a list of all mailboxes.


The Get-MSOLPartnerContract looks interesting, lets try that.


Well, that gets me a lot of Tenant ID’s, which doesn’t appear to be useful. Or does it? I know from using Office 365 with PowerShell in the past that there are commands that use the TenantID parameter to retrieve information.

Doing some more investigation I run the series of commands below. First I am storing all the Tenant ID’s in a $Clients variable, and then using Get-Member to see what properties this variable contains. Seeing that one of them is Default Domain Name, I run $Clients.DefaultDomainName which sure enough enumerates all of my client domain names so I can see that they are in fact my clients. I am then storing all the Tenant ID’s in another variable $Tenants which I will be using shortly.

Earlier I mentioned that I knew from past experience that there were cmdlets in the MSOnline module that used TenantID as a parameter. How could I find out what those commands are? By running the command below. I reduced the output by using -Verb Get since I am going to be getting the number of mailboxes.

Scrolling through the list I see that Get-MSOLUser is one of the cmdlets that uses TenantID as a parameter and think that might be a good place to start. I take one of my Tenant ID’s in my $Tenants variable and run the command below.

I was hoping to see that there was a domain or client name property, but unfortunately there is not. So in order to figure that out I ran the commands below. What I am doing here is taking the first user that comes up for the tenant and getting their User Principal Name (UPN). Since each client’s users have the same UPN, this will be OK. After that I am splitting the UPN at the ‘@’ symbol and using that information for the domain. This is definitely not the only way to do this (or probably the best) but it worked for me.

Next I am creating a hashtable with my properties and then creating a new PSObject Object and saving this information to a .CSV file. Notice in the properties section that for the Domain I am having to use $Domain[1] because when it split the UPN at the ‘@’ symbol it split it into two pieces, the section before the ‘@’ and the section after it.

Finally, I have tied all this into a Function I am calling Get-PartnerMailboxes that will iterate through all my Tenant ID’s and create a .CSV listing the domain and total number of licensed users for each tenant.

Desired State Configuration: Cannot bind argument to parameter ‘ResourceKey’ because it is an empty string

Applies To: PowerShell Version: 5.0.10240.16384, xComputerManagement 1.2.2, 1.3.0

I was working on a DSC Configuration to build a Hyper-V host, when I encountered the error shown below.

So, that’s a pretty awesome error, and not one that I can say that I have seen before. I did A LOT of troubleshooting to figure out what was going on (which in the interests of time and space I am going to skip over) and we will get right to the good stuff.

My Configuration had some string parameters that were not assigned any values in the Configuration, or when I was creating the .mof files. Let’s pretend the Configuration looks like the one below. If you just copy and paste and run this, it is going to fail with the error above.

If you remove the section for xComputerManagement, it will work just fine. Also, if you hard code a value into the parameter $ComputerName in the Configuration, or pass it in when you build the Configuration (I have shown both ways below), it will also not error and will work the way you would expect it to.

So, the next question is, why is this? Thankfully, the answer is in the error message once you figure it out. The Name parameter of the DSC resource is the Key value of the resource, and as such it cannot be an empty string (which completely makes sense). It doesn’t throw the error for the File resource, because it’s key value is DestinationPath, which in my example I have coded in as an actual value. How can you tell what the Key property of a DSC Resource is? Look at the example below! Key properties are Mandatory parameters, and as such are shown without being enclosed in square brackets. You can also open up and look at the schema.mof file for the resource which will show you that information as well.

To further illustrate the point, if I run the Configuration below, it is also going to error out.

Warning: Unable to find package provider ‘PSModule’ on Windows 10

Running Windows 10 Enterprise Preview Build 5.0.10158.0

When I run Find-Module, I receive the error message in the screenshot below.

After bashing my face against the problem for a while, I got some help from Ben Gelens, who provided the answer on how to fix the problem:

Once that is done, close and reopen the ISE (or the Console host) and you are good to go!

Unexpected Token ‘}’ in Expression or Statement when using $AllNodes in a Desired State Configuration

I was working on testing a custom DSC Resource when I came across a very strange issue. Whenever I tried to build a Configuration like the one below, it failed.

Under the very last curly brace there was a red squiggly line. When I tried to build the Configuration it failed with the error:

It took me nearly an hour and a half to figure out what was causing the issue. Here is the working Configuration. Can you spot the difference?

If you can’t spot the difference, the only difference is that in the second example, there is a space between $AllNodes.NodeName and the curly brace beginning the Node block. As you can imagine I have a serious problem with this, and believe this to be a bug. You will notice that I don’t need a space between the name of the Configuration and the curly brace, so why should I need a space after $AllNodes.NodeName? Also, in the next example, if I change $AllNodes.NodeName to Node MYPC, it doesn’t require a space between the node name and the curly brace either.

All of this testing was done on Windows 10 Enterprise Preview edition with PowerShell Version 5.0.10158.0 . If you also believe this is a bug and should be fixed, vote up the issue on connect here.

Why Should I Attend the PowerShell Summit?

Starting last fall for the European PowerShell Summit and this year for the North America PowerShell Summit all the session recordings are available on YouTube. Because of that you may be asking yourself “Why should I go to the Summit when I can just watch everything online?”. Here are ten reasons, in no particular order other than the order in which my brain dumped them out.

  1. You get direct interaction with the product team. And I am not talking about members of the PowerShell team sitting in a corner not interacting with anyone.  They are there to get as much feedback as possible, learn how people are using it, and to directly interact with members of the community.  This isn’t the Microsoft of 10 years ago or two years ago.  When they say they want your feedback (and not just the good stuff) they absolutely mean it.  Special thanks to Lee Holmes, Michael Greene, Joey Aiello, Angel Calvo, Hemant Mahawar, and Kenneth Hansen for making the trip out to Charlotte.
  2. I don’t care how much you interact with other members of the community over Twitter, Email, Google Hangouts, whatever.  There is no substitute for meeting people face to face, shaking their hand and getting to know something about them besides how they use PowerShell.
  3. And combining #1 and #2, you also get to talk to (and listen to) people talk about how they have solved problems using PowerShell, and what their thought process was around creating that solution. You can then ask people, I have this problem, what would you do to solve it?  One of those conversations alone is worth the price of admission.
  4. You get to watch Mike Robbins “harass” Rohn Edwards all week by telling everyone how great his sessions are going to be and how everyone needs to go to them.  By all accounts they were awesome.
  5. You get to see the look on Dave Wyatt’s face Microsoft announces that “his code” (Pester) is shipping with the next version of Windows Server
  6. You get to have Steven Murawski answer your questions about creating Custom DSC Resources while you are creating them
  7. You get a free Chef T-Shirt, courtesy of Steve.
  8. You get awesome Nano Server and PowerShell stickers courtesy of the one and only Jeffrey Snover.
  9. You get to watch Jason Helmick live and in person talk about how he has his depends on.
  10. You end up finding a bug in Class based DSC Resources that you only found because you participated in the DSC Resource hack-a-thon at the PowerShell Summit.  So make sure you vote on that!
  11. Bonus!  Jeff Hicks gives you a signed PowerShell Deep Dives book and a 30 day Pluralsight subscription to give away at your next user group meeting.
  12. Bonus!  You get to watch Jeffrey Snover demonstrate and talk about a bunch of stuff I can’t repeat or talk about upon fear of death :).
  13. Bonus!  You get to talk to (and listen) to June Blender talk about PowerShell, PowerShell Help, and writing.  Her passion and knowledge around those topics is unbelievable.
  14. Bonus!  You learn how little you really know about PowerShell.  This is a good thing!  This is also something I relearn on a nearly daily basis.

I was also asked by Josh Duffney on Twitter what I thought were some of the “must watch” videos from the Summit.  The lame answer is “everything”, but that’s also not realistic.  If you put a gun to my head and said “you have to pick 7 sessions” here are the 7 I would pick (no particular order).  All the Summit videos can be found in this playlist on YouTube.

  1. Kenneth Hansen & Angel Calvo PowerShell Team Engagement
  2. Don Jones DSC Resource Design
  3. Dave Wyatt on Automated Testing using Pester
  4. Defending the Defenders Part 1 & 2
  5. Debugging
  6. PowerShell Get
  7. Ashley McGlone on DSC and AD
  8. PowerShell v5 Debugging (There is also a session on Debugging PowerShell by Kirk Munro.  These are different)

ValidateSet for a Parameter in a DSC Class Based Resource Fails to Throw Error

While working on a Custom DSC Resource that I started Monday night at the PowerShell Summit I came across some interesting behavior that turned out to be a bug in the WMF 5.0 February Preview. I have logged this issue in Connect, but I wanted to write a blog post to demonstrate what exactly is going on for when someone else runs into this issue. I am just going to the use Custom DSC Resource for creating a Primary DNS Zone that I was working on as the example to demonstrate the behavior.

Here are the properties for the resource. I figured I could just do ValidateSet like I always had done for an advanced function or non-class based DSC Resource.

My Configuration for testing the Resource looks like this:

If I run this Configuration, with one of the appropriate values for ReplicationScope, it works exactly like you would expect it to.

That’s great. But, what happens if I put in a value that isn’t part of Validate Set?

That is clearly not what should happen. You would expect to see an error saying something to effect of “HokeyPokey does not belong to the set “Domain”,”Forest”,”Current”,”Legacy”, it needs to be one of those values”.

If we look at the .MOF file that gets created, this incorrect value also makes it into the .MOF file:

You can tell that it knows something is wrong, because when it runs through Test-TargetResource and Set-TargetResource it doesn’t actually do anything (notice all the Verbose messages that are missing from the previous example), but it also doesn’t error.

So how do we get around this? By using an Enum!

Now, if I try to set the ReplicationScope to HokeyPokey, we get the behavior we would expect.

Creating a Class based DSC Resource using PowerShell

Scenario: I would like to think I am fairly competent at creating DSC Resources using .MOF files. But with the upcoming release of Windows Management Framework 5.0 we can now write DSC Resources using Classes. I have never done anything with a Class. I am going to attempt to figure out how to write a simple Class based DSC Resource.

I am going to start with this TechNet Article, because it was the first thing that came up when I Googled “Create Class based DSC Resource”.

I am going to create a DSC Resource called MyTestClassResource. The first thing I need to do is create the appropriate folder structure, which I do by running the following two commands:

Next, I need to create the Module Manifest.

The .PSM1 file is where I define and create my Class based resource.

Now, let’s get to creating this Resource. I am going to go super simple here and create a Resource that will just ensure that a Folder exists. Yes, I know the File DSC Resource already does this, that’s not the point :). After fiddling around for a little bit here is what my Class DSC Resource “Skeleton” looks like.

The rest of this should be pretty straightforward. Right? The first thing I try to do in the Get section is to test and see if the path exists.

This doesn’t work however because “Variable is not assigned in the method”, whatever the hell that means. Looking at the article it uses a $This object (I have no idea if that’s even the right word) with the variables to do things, so let’s try a different tactic.

And that works fine, so clearly $This is some special thing I need to be using going forward. This should be fun :). Now that I got that working, the rest of this Method was pretty straight forward.

After that, the Test Method is pretty easy as well. However, I have no idea what is going on with the whole Return -not $Item part, I am just following along from the example and hoping that it works.

Onto the Set Method. One thing I noticed when creating this is that I don’t need an Else block with an If statement, which is nice.

With that done I need to re-create the Module Manifest that I made earlier with some important information.

At this point while trying to run Get-DSCResource I realized that my folder structure I created at the beginning was not correct, because I was used to the way I had been doing it when working with .MOF Files. I actually need only this:

And then I moved the .psm1 and .psd1 files into that folder. No extra sub-folder required. This is a win in my book.

Now, when I run Get-Module -ListAvailable, the MyTestClassResource Module is listed. However, when I run Get-DSCResource -Module MyTestClassResource I get a lot of nothing. Weird. Next I try to Import-Module -Name MyTestClassResource and I get this giant bundle of joy.

Uhhh…What? Thinking quickly I decide that it’s probably not going to work to name everything exactly the same. I rename the MyTestClassResource folder to MyTestClassResources and leave everything else the same. And well, I will spare all the error text but that didn’t work at all either. I am not sure how much time I spent trying to figure out what the hell I was doing wrong, but to say it was an exercise in frustration is a massive understatement. No matter what I did I couldn’t get it to recognize a valid Module, let alone DSC Resource. I lost track of all the troubleshooting that I did but here is what my folder structure looks like now that it is working. Note that I renamed my Class based Resource from MyTestClassResource to MyFolderResource.

Now, let’s test a Configuration!

And just to make sure it’s working, let’s set it to Absent

cVSS Custom Desired State Configuration (DSC) Resource

I will keep this pretty short and sweet. I was asked to create a Custom DSC Resource that could ensure that Volume Shadow Copies was always enabled on a drive. I thought it would be pretty easy to do, but turns out it wasn’t nearly as easy as I thought. When you enable VSS through the GUI there are two things that happen. First, VSS gets enabled on the drive with some default settings. Second, two scheduled tasks are created that match those settings to actually create the shadow copies. Because of that I broke this custom resource into two parts. The cVSS Resource just specifies the drive you want to enable, whether or not you want to enable it, what drive you want to store the shadow copies on, and how much space you want the shadow copies to take up. A Configuration using this Resource looks like this:

The cVSSTaskScheduler Resource creates the Scheduled Tasks that create the actual shadow copies on the drive. A Configuration using this Resource looks like this:

All the files you need to use this Resource can be found on my GitHub site. There is an Examples folder that has the two examples above plus a Configuration using both. The Files folder contains two files that just show how I created (and Updated) the Resource as I went through the process of actually getting it to work. I am very new to the whole GitHub thing so if there are bugs, things that I missed, or a better way of doing something and you want to update or change the files let me know and we can certainly do that.