Query for 32-bit or 64-bit Versions of Microsoft Office with ConfigMgr

Quick post – I needed to query for 64-bit versions of Microsoft Office installed on our clients. Usually, the 32-bit version gets installed as this is Microsoft’s recommendation due to add-in compatibility etc. But in some cases, the 64-bit version is required to take advantage of additional RAM. I couldn’t find anything useful online to distinguish between the x86 and x64 versions of Office using inventory data from ConfigMgr, but I found that the product code GUID of the Office product actually contains this information (see here).

So I put together the following SQL query which extracts some additional data about Office from the GUID, including the Release Version (ie RTM, SP1 etc), Release Type (ie Volume License, Trial etc) and the Bit Version.

This should work for any version of Office, but only for the MSI installer (ie not Click-to-Run).

Some example results:

SQLQuery


Select
  sys.Name0 as 'Device Name',
  sof.ProductName0 as 'Product Name',
  sof.ProductVersion0 as 'Product Version',
  sof.InstallDate0 as 'Installation Date',
  sof.Language0 as 'Language Code',
  sof.SoftwareCode0 as 'Software Code',
  'Release Version' =
    Case substring(sof.SoftwareCode0,2,1)
      When '0' Then 'Prior to Beta 1'
      When '1' Then 'Beta 1'
      When '2' Then 'Beta 2'
      When '3' Then 'RC0'
      When '4' Then 'RC1 / OEM Preview'
      When '9' Then 'RTM'
      When 'A' Then 'SP1'
      When 'B' Then 'SP2'
      When 'C' Then 'SP3'
      Else 'Unknown'
    End,
  'Release Type' =
    Case substring(sof.SoftwareCode0,3,1)
      When '0' Then 'Volume License'
      When '1' Then 'Retail / OEM'
      When '2' Then 'Trial'
      When '5' Then 'Download'
      Else 'Unknown'
    End,
  substring(sof.SoftwareCode0,4,2) as 'Major Version',
  substring(sof.SoftwareCode0,6,4) as 'Minor Version',
  'Bit Version' =
    Case substring(sof.SoftwareCode0,21,1)
      When '0' Then '32-bit'
      When '1' Then '64-bit'
      Else 'Unknown'
    End
from v_GS_INSTALLED_SOFTWARE sof
inner join v_R_System sys on sof.ResourceID = sys.ResourceID
where sof.SoftwareCode0 like '%0ff1ce%'
-- Querying for Office Professional Plus --
and sof.ProductName0 like '%Professional Plus%'
-- Querying for 64-bit Office (0 = x86, 1 = x64) --
and substring(sof.SoftwareCode0,21,1) = 1
Order by sys.Name0 

Inventory Local Administrator Privileges with PowerShell and ConfigMgr

Any security-conscious enterprise will want to have visibility of which users have local administrator privilege on any given system, and if you are an SCCM administrator then the job of gathering this information will likely be handed to you!

However, this task may not be as simple as it seems. Gathering the membership of the local administrators group is one thing, but perhaps more important to know is whether the primary user of a system has administrator privileges. If that user is a member of a group that has been added to the local administrators group, then it isn’t immediately obvious whether they actually have administrator rights without also checking the membership of that group. And what if there are further nested groups – ie the user is a member of a group that’s a member of a group that’s a member of the local administrators group?! Obviously things can get complicated here, making reporting and compliance checking a challenge.

Thankfully, PowerShell can handle complication quite nicely, and ConfigMgr is more than capable as a both a delivery vehicle and a reporting mechanism, so the good news is – we can do this!

The following solution uses PowerShell to gather local administrator information and stamp it to the local registry. A Compliance item in SCCM is used as the delivery vehicle for the script and then RegKeyToMof is used to update the hardware inventory classes in SCCM to gather this information from the client’s registry into the SCCM database, where we can query and report on it.

Gathering Local Administrator Information with PowerShell

To start with, let’s have a look at some of the PowerShell code and the information we will gather with it.

First, we need to identify who is the primary user of the system. Since the script is running locally on the client computer, we will not use User Device Affinity. True, UDA information is stored in WMI in the CCM_UserAffinity class, in the ROOT\CCM\Policy\Machine\ActualConfig namespace.  But this class can contain multiple instances so you can’t always determine the primary user that way.

A better way is to use the SMS_SystemConsoleUsage class in the ROOT\cimv2\sms namespace and query the TopConsoleUser property. This will give you the user account who has had the most interactive logons on the system and for the most part will indicate who the primary user is.


$TopConsoleUser = Get-WmiObject -Namespace ROOT\cimv2\sms -Class SMS_SystemConsoleUsage -Property TopConsoleUser -ErrorAction Stop | Select -ExpandProperty TopConsoleUser

Next, to find if the user is a local admin or not, we will not simply query the local administrator group membership and check if the user is in there. Instead we will create a WindowsIdentity object in .Net and run a method called HasClaim(). I describe this more in a previous blog, but using this method we can determine if the user has local administrator privilege whether through direct membership or through a nested group.


$ID = New-Object Security.Principal.WindowsIdentity -ArgumentList $TopConsoleUser
$IsLocalAdmin = $ID.HasClaim('http://schemas.microsoft.com/ws/2008/06/identity/claims/groupsid','S-1-5-32-544')
$ID.Dispose()

The SID for the local admin group (S-1-5-32-544) is used as this is the same across all systems. This will only work for domain accounts as it uses kerberos to create the identity.

Now we will also get the local administrator group membership using the following code (more .Net stuff), and filter just the SamAccountNames.


Add-Type -AssemblyName System.DirectoryServices.AccountManagement -ErrorAction Stop
$ContextType = [System.DirectoryServices.AccountManagement.ContextType]::Machine
$PrincipalContext = New-Object -TypeName System.DirectoryServices.AccountManagement.PrincipalContext -ArgumentList $ContextType, $($env:COMPUTERNAME) -ErrorAction Stop
$IdentityType = [System.DirectoryServices.AccountManagement.IdentityType]::Name
$GroupPrincipal = [System.DirectoryServices.AccountManagement.GroupPrincipal]::FindByIdentity($PrincipalContext, $IdentityType, “Administrators”)
$LocalAdminMembers = $GroupPrincipal.Members | select -ExpandProperty SamAccountName | Sort-Object
$PrincipalContext.Dispose()
$GroupPrincipal.Dispose()

Next, if the user is a local admin through nested group membership, I will call a custom function which will check the nested group membership within the local admin group, for the user account. Let’s say that Group B is a member of Group A, which is a member of the local administrators group. We will check the membership of both Groups B and A to see which ones the user is a member of, and therefore which group/s is effectively giving the user administrator privilege. We do this by querying the $GroupPrincipal object created in the previous code. The custom function will query nested membership up to 3 levels deep.

Now I will query the Install Date for the operating system, since in some cases where a machine is newly built, the TopConsoleUser may not yet be the primary user of the system, but the admin who built the machine, for example. This date helps to identify any such systems.


[datetime]$InstallDate = [System.Management.ManagementDateTimeConverter]::ToDateTime($(Get-WmiObject win32_OperatingSystem -Property InstallDate -ErrorAction Stop | Select -ExpandProperty InstallDate)) | Get-date -Format 'yyyy-MM-dd HH:mm:ss'

Now we gather all this information into a datatable, and call another custom function to write it to the local registry. I use the following registry key, but you can change this in the script if you wish:

HKLM:SOFTWARE\IT_Local\LocalAdminInfo

The script will create the key if it doesn’t exist.

Here’s an example of the kind of data that will be gathered:

localadmin

You can see in this example, that my user account is a local administrator both by direct membership and through nested groups. The actual groups that grant this right are listed in the NestedGroupMembership property.

Create a Compliance Item

Now lets go ahead and create a compliance item in SCCM to run this script.

In the Console, navigate Assets and Compliance > Compliance Settings > Configuration Items.

Click Create Configuration Item

config1

Click Next and select which OS’s you will target.  Remember the Windows XP and Server 2003 may not have PowerShell installed.

Click Next again, then click New to create a new setting.

Choose Script as the setting type, and String as the data type.

config2

Now we need to add the scripts.  You can download both the discovery and remediation scripts from my Github repo here:

https://github.com/SMSAgentSoftware/ConfigMgr/tree/master/PowerShell%20Scripts/Compliance%20Settings/LocalAdministratorInfo

Click Add Script and paste or open the relevant script for each. Make sure Windows Powershell is selected as the script language.

The discovery script simply checks whether the script has been run in the last 15 minutes, and if not returns non-compliant.  This allows the script to run according to the schedule you define for it, ie once a day or once a week etc, to keep the information up-to-date in the registry.

The remediation script does the hard work 🙂

Click OK to close the Create Setting window.

Click Next, then click New to create a new Compliance Rule as follows:

config3

Click OK to close, then Next, Next and Close to finish.

Create a Configuration Baseline

Click on Configuration Baselines and Create Configuration Baseline to create a new baseline.

Give it a name, click Add and add the Configuration Item you just created.

config4

Click OK to close.

Deploy the Baseline

Right-click the baseline and choose Deploy. Make sure to remediate noncompliance and select the collection you wish to target.

config5

Update SCCM Hardware Inventory

Creating the MOF Files

For this part you will need the excellent RegKeyToMOF utility, which you can download from here:

https://gallery.technet.microsoft.com/RegKeyToMof-28e84c28

You will also need to do this on a machine that has either run the remediation script to create the registry keys, or has run the configuration baseline.

Open RegKeyToMOF and browse to the registry key:

HKLM:SOFTWARE\IT_Local\LocalAdminInfo

You can deselect the ‘Enable 64bits …’ option as the registry key is not located in the WOW6432Node.

Click Save MOF to save the required files.

regkey

Copy the SMSDEF.mof and the CM12Import.mof to your SCCM site server.

Update Client Settings

In the SCCM console, navigate Administration > Site Configuration > Client Settings. Open your default client settings and go to the Hardware Inventory page.

Click Set Classes…, then Import…

Browse to the CM12Import.mof and click Import.

import

Close the Client Settings windows.

Update Configuration.mof

Now open your configuration.mof file at <ConfigMgr Installation Directory> \inboxes\clifiles.src\hinv.

In the section at the bottom for adding extensions, which starts like this…

//========================
// Added extensions start
//========================

…paste the contents of the SMSDEF.mof file.  Save and close the file.

Reporting

Now that you’ve deployed the configuration item and updated the SCCM hardware inventory, a new view called dbo.v_GS_LocalAdminInfo0 has been added to the SCCM database. Note that initially there will be no data here until your clients have updated their policies, ran the configuration baseline, and ran the hardware inventory cycle.

You can query using the Queries node in the SCCM console…

query

…or create yourself a custom SCCM report, create an Excel report with a SQL data connection, query the SCCM database with PowerShell – whatever method you need or prefer.

Here is a sample SQL query that will query the view and add some client health data and the chassis type to help distinguish between desktop, laptops, servers etc.


Select
  ComputerName0 as 'ComputerName',
  Case When enc.ChassisTypes0 = 1 then 'Other'
    when enc.ChassisTypes0 = 2 then 'Unknown'
    when enc.ChassisTypes0 = 3 then 'Desktop'
    when enc.ChassisTypes0 = 4 then 'Low Profile Desktop'
    when enc.ChassisTypes0 = 5 then 'Pizza Box'
    when enc.ChassisTypes0 = 6 then 'Mini Tower'
    when enc.ChassisTypes0 = 7 then 'Tower'
    when enc.ChassisTypes0 = 8 then 'Portable'
    when enc.ChassisTypes0 = 9 then 'Laptop'
    when enc.ChassisTypes0 = 10 then 'Notebook'
    when enc.ChassisTypes0 = 11 then 'Hand Held'
    when enc.ChassisTypes0 = 12 then 'Docking Station'
    when enc.ChassisTypes0 = 13 then 'All in One'
    when enc.ChassisTypes0 = 14 then 'Sub Notebook'
    when enc.ChassisTypes0 = 15 then 'Space-Saving'
    when enc.ChassisTypes0 = 16 then 'Lunch Box'
    when enc.ChassisTypes0 = 17 then 'Main System Chassis'
    when enc.ChassisTypes0 = 18 then 'Expansion Chassis'
    when enc.ChassisTypes0 = 19 then 'SubChassis'
    when enc.ChassisTypes0 = 20 then 'Bus Expansion Chassis'
    when enc.ChassisTypes0 = 21 then 'Peripheral Chassis'
    when enc.ChassisTypes0 = 22 then 'Storage Chassis'
    when enc.ChassisTypes0 = 23 then 'Rack Mount Chassis'
    when enc.ChassisTypes0 = 24 then 'Sealed-Case PC'
    else 'Unknown'
  End as 'Chassis Type',
  TopConsoleUser0 as 'Primary User',
  TopConsoleUserIsAdmin0 as 'Primary User is Admin?',
  AdminGroupMembershipType0 as 'Primary User Local Admin Group Membership Type',
  LocalAdminGroupMembership0 as 'Local Admin Group Membership',
  NestedGroupMembership0 as 'Primary User Local Admin Nested Group Membership',
  OSAgeInDays0 as 'OS Age (days)',
  OSInstallDate0 as 'OS Installation Date',
  LastUpdated0 as 'Last Updated Date',
  la.TimeStamp as 'HW Inventory Date',
  ch.ClientStateDescription,
  ch.LastActiveTime
from dbo.v_GS_LocalAdminInfo0 la
join v_R_System sys on la.ComputerName0 = sys.Name0
left join v_GS_SYSTEM_ENCLOSURE enc on sys.ResourceID = enc.ResourceID
left join v_CH_ClientSummary ch on sys.ResourceID = ch.ResourceID
where ComputerName0 is not null
  and enc.ChassisTypes0 <> 12

 

PowerShell Custom Class for Querying a SQL Server

Here is a handy custom class I created for PowerShell 5+ that can query an SQL database. The class creates a SQL connection using the System.Data.SqlClient.SqlConnection class in .Net, and gives you full access to this object for information or to manipulate parameters. The class can also query the list of tables and views available in the database to help you prepare your SQL query.

Here’s a quick how-to:

Instantiate the Class

Create an instance of the object. You can create an empty instance like this:


$SQLQuery = [SQLQuery]::new()

You can then specify the SQL Server name and Database name:


$SQLQuery.SQLServer = "SQL-01\Inst_SCCM"
$SQLQuery.Database = "CM_ABC"

And then add a query:


$SQLQuery.Query = "Select top 5 * from v_R_System"

Or you can pass these parameters when you create the object, like:


$SQLQuery = [SQLQuery]::new("SQL-01\Inst_SCCM","CM_ABC")

Or..


$SQLQuery = [SQLQuery]::new("SQL-01\Inst_SCCM","CM_ABC","Select top 5 * from v_R_System")

Execute the Query

Now execute the query to return the results:


$SQLQuery.Execute()

The result will be displayed in the console window, or you can pipe to GridView for easier viewing:


$SQLQuery.Execute() | Out-GridView

Grid1

The query results will be stored to the object in the Result parameter so you can retrieve them later.


$SQLQuery.Result

The results are stored in a DataTable format with rows and columns which is an ideal format for SQL query results.

Load a Query from a File

You can load in a SQL query from a “.sql” file like so, passing the location of the file to the method:


$SQLQuery.LoadQueryFromFile("C:\Scripts\SQLScripts\OSD_info.sql")

Then execute the query as before.

Change the Connection String

By default, the connection string will be created for you, but you can add your own custom connection string using the ConnectionString parameter:


$SQLQuery.ConnectionString = "Server=SQL-01\inst_sccm;Database=CM_ABC;Integrated Security=SSPI;Connection Timeout=5"

Timeout Values

There are two timeout parameters which can be set:


$SQLQuery.ConnectionTimeout
$SQLQuery.CommandTimeout

The ConnectionTimeout parameter is the maximum time in seconds PowerShell will try to open a connection to the SQL server.

The CommandTimeout parameter is the maximum time in seconds PowerShell will wait for the SQL query to execute.

Get a List of Views or Tables in the Database

The following two methods can be used to retrieve a list of views or tables in the database:


$SQLQuery.ListViews()
$SQLQuery.ListTables()

The views or tables will be stored to the parameter of the same name in the object, so you can retrieve them again later.

You can filter the list to search for a particular view or table, or group of. On the command line you could do:


$SQLQuery.ListViews()
$SQLQuery.Views.Rows | where {$_.Name -match "v_Client"}

Or as a one-liner:


$SQLQuery.ListViews().Where({$_.Name -match "v_Client"})

You could also output to GridView for filtering:


$SQLQuery.ListViews() | Out-GridView

Grid2

Hide the Results

By default, the Execute(), ListViews() and ListTables() methods will return the results to the console after execution. You can turn this off by setting the DisplayResults parameter to $False. This scenario may be useful for scripting where you may not wish to display the results right away but simply have them available in a variable.


$SQLQuery.DisplayResults = $False

Access the SQL Objects

You can access the SQLConnection object, or the SQLCommand object to view or set parameters, or execute any of the methods contained in those objects. These objects are only available once the SQL query has been executed.

SQLConnection

SQLCommand

You can re-use the same SQLQuery object to run other queries; only the results for the most recent query will be stored in the object itself.

SQLQuery Custom Class

Export / Backup Compliance Setting Scripts with PowerShell

In my SCCM environment I have a number of Compliance Settings that use custom scripts for discovery and remediation, and recently it dawned on me that a lot of time has been spent on these and it would be good to create a backup of those scripts. It would also be useful to be able to export the scripts so they could be edited and tested before being updated in the Configuration Item itself. So I put to together this PowerShell script which does just that!

The Configuration Item scripts are stored in an XML definition, and this can be read from the SCCM database directly and parsed with PowerShell, so that’s what this script does. It will load all the Configuration Items into a datatable from a SQL query, then go through each one looking for any settings that have scripts defined. These scripts will be exported in their native file format.

You could then edit those scripts, or add the export location to your file/folder backup for an extra level of protection for your hard work!

Here you can see an example of the output for my “Java Settings” Configuration item. A subdirectory is created for the current package version, then subdirectories under that for each Configuration setting, then the discovery and remediation scripts for that setting.

cis
Exported Configuration Item Scripts

Note that the script will only process Compliance Items with a CIType_ID of 3, which equates to the Operating System type you will see in the SCCM console for the Configuration Item, which is the type that may use a script as the discovery source.

Export-CMConfigurationItemScripts.ps1

Automatically Set SQL MaxServerMemory on Cluster Failover with PowerShell

On a two-node Windows Failover Cluster, I have two SQL instances installed. Each instance runs on its own node in the cluster, so that it can make maximum use of the available memory on that server. However, when a failover occurs, it is necessary to reduce the maximum server memory setting for both instances, so that they can share the available memory on the one server. Rather than have to do that manually, however, I decided to automate the process using PowerShell and the Windows Task Scheduler, and here’s how.

Note: the script will work for two SQL instances in a cluster. For additional instances, the script will need to be updated accordingly.

  1. Save the PowerShell script (download from Technet Gallery) to each server in the cluster.
  2. Update the PowerShell script setting the required variables in the parameters section, such as the log file location, the SQL instance names, the SQL service names, the path to the SQL SMO dll, the maximum server memory limit you want to set, and the timeout period.
  3. Create a scheduled task on each server, running as an account that has the appropriate permissions on each instance. Add 2 triggers – one for each SQL instance – and use the event trigger. You can use event ID 101, for example, to identify when a SQL instance becomes active on that node. As the source, use the “SQLAgent$<instancename>”.

event

As the task action, use Start a program:

  • Program: Powershell.exe
  • Arguments: -ExecutionPolicy Bypass -WindowStyle Hidden -File “<PathToScript>\Set-SQLClusterMaximumMemory.ps1”

When a SQL instance fails over to the other node, the script will be triggered and will set the maximum server memory limit for both instances on that node.

maxmem

The script will also log the process:

log

Automation. Gotta love it 🙂