Skip to main content

Reading and Resolving PowerShell Errors - Part 10

This is part 10 of my series of the most common PowerShell errors that are made in my PowerShell classes.  This is the last one!
Today’s error: My data file does not have my data.

Here is our starting code:
Get-Process |
    Measure-Object -Property CPU -sum |
    Select-Object -Property Name, CPU |
    Export-CSV -Path c:\temp\data.csv

And here is the full error:
No error, just the wrong data.  Here is the contents of the CSV.
PS C:\temp> import-csv -Path C:\temp\Data.csv

Name CPU
---- ---

Resolution:
The problem here is that the programmer did not keep track of what is in the PowerShell Pipeline. I see this very often.  When you are constructing a pipeline, you need to keep track of what is in the pipeline.  After each statement, pipe to object to Get-Member and verify that you are still working with the object type that you expect.
PS C:\temp> Get-Process | Get-Member


   TypeName: System.Diagnostics.Process

Here you can see that the original object is System.Diagnostics.Process.  Now look at the member information after the execution of the second command.
PS C:\temp> Get-Process |
    Measure-Object -Property CPU -sum | Get-Member


   TypeName: Microsoft.PowerShell.Commands.GenericMeasureInfo

Name        MemberType Definition                               
----        ---------- ----------                               
Equals      Method     bool Equals(System.Object obj)           
GetHashCode Method     int GetHashCode()                        
GetType     Method     type GetType()                           
ToString    Method     string ToString()                        
Average     Property   System.Nullable[double] Average {get;set;}
Count       Property   int Count {get;set;}                     
Maximum     Property   System.Nullable[double] Maximum {get;set;}
Minimum     Property   System.Nullable[double] Minimum {get;set;}
Property    Property   string Property {get;set;}               
Sum         Property   System.Nullable[double] Sum {get;set;}

The object typename is now Microsoft.PowerShell.Commands.GenericMeasureInfo.  In other words, you have a completely new object.  Many cmdlets alter the object in the pipeline. The next command, Select-Object will attempt to filter out all but two properties.  The problem here is that the properties that are being asked for are from the object System.Diagnostics.Process, not the current object in the pipeline.
To fix this, remove the Measure-Object cmdlet.
Get-Process |
    Select-Object -Property Name, CPU |
    Export-CSV -Path c:\temp\data.csv


This mistake is most often the result of not wanting to make the effort.  Once you are experienced working with a set of cmdlets, you will know what they put into the pipeline.  Until you are experienced with a set of cmdlets, make sure you check the pipeline often.  This will save you a lot of work in debugging.

Comments

Popular posts from this blog

With the AD Recycle Bin Turned on, What Happens when you Create a User Account with a Password that does not meet the Password Policy?

This was an interesting observation from one of my Windows Server 2012 classes.  While working with the AD Recycle bin in a lab, one of my students discovered some interesting accounts that were created. When he created user accounts that did not meet password complexity requirements, an account is temporarily made and then deleted.  When a new password is provided that meets the password requirements, then a new account is made. We discovered this in two places.  First off in the Active Directory Administrative Center.  This is what caused the initial confusion.  Take a look.  This is in the Deleted Objects OU. You can see multiple deleted accounts for Test2 and one for Test3.  Test3 is a valid, functioning user account.  Using the PowerShell command Get-ADObject –IncludeDeletedObjects –Filter * –Properties ObjectSID we can see that indeed, two accounts were created, with one of them deleted. Notice the RID portion of the SID is different. ...

Sticky Key problem between Windows Server 2012 and LogMeIn

This week I instructed my first class using Windows Server 2012 accessed via LogMeIn and discovered a Sticky Key problem every time you press the Shift key. Here is my solution to resolve this.  First off, in the Preferences of LogMeIn for the connection to the Windows Server, click General . Change the Keyboard and mouse priority to Host side user and click Apply at the bottom. On the Windows 2012 server, open the Control Panel – Ease of Access – Change how your keyboard works . Uncheck Turn on Sticky Keys . Click Set up Sticky Keys . Uncheck Turn on Sticky Keys when SHIFT is pressed five times . Click OK twice. If you are using Windows Server 2012 as a Hyper-V host, you will need to redo the Easy of Use settings on each guest operating system in order to avoid the Sticky Key Problem. Updated Information: March 20, 2013 If you continue to have problems, Uncheck Turn on Filter Keys .

Backup and Restore AD LDS with DSDBUTIL.exe

Active Directory Lightweight Directory Services allow you to create a directory service that allows applications to have access to user accounts, groups, and authentication similar to Active Directory Domain Services.  The big advantage here is that the schema of the directory service will not be bound by the rules of an Active Directory database.  Exchange 2007/2010, for example, use an instance of AD LDS on the Edge Transport Server to provide for user authentication from the internet.  Because your Active Directory database is not exposed to the internet, this is more secure. Applications will handle most of the dirty work should they require AD LDS.  You may want to make sure the database is being backed up and also have a restore plan in place.  Should the database become corrupt, the application that uses that database will fail.  This document will walk you through backing up and restoring an instance of AD LDS using the dsdbutil.exe command. Fi...