| Downloads |
|---|
| 97609.zip |
| Executive Summary: Kb.ps1 is Windows PowerShell script that uses Windows Management Instrumentation (WMI) to check whether a specific patch is installed on computers and provides the results in an onscreen report. To make it easier to spot problematic computers, this Windows PowerShell script displays the names of the computers that aren't patched in red and the names of the computers that are successfully patched in green. |
Each time Microsoft Software Update Services (SUS) downloads patches, there are always a few machines not properly patched due to unforeseeable problems. Because SUS doesn’t have any reporting tools, it’s difficult to determine which machines aren’t patched. Thus, administrators at my company have been using KB.vbs to report the status of patches on more than 600 Windows client machines. In my Reader to Reader article “Does SUS Make You Want to Send an SOS?” (www.windowsitpro.com/Articles/ArticleID/46953/46953.html), I present this VBScript script, which I wrote.
Although KB.vbs works well, I decided to rewrite the script in PowerShell last July because I wanted to experience the power in PowerShell firsthand. More important, I never want to stop learning. I feel it’s important to learn to apply PowerShell to current problems.
The result is kb.ps1. Like KB.vbs, kb.ps1 first attempts to ping the machines listed in an input file named pclist.txt. If a PC is online, the script determines whether the specified patch exists and reports the results. To make it easier to spot problematic computers, the names of the computers that aren’t patched are displayed in red and the names of the computers that are successfully patched are displayed in green. The script also reports ping failures.
Figure 1 shows a sample input file. The path to this input file and the patch to search for are specified on the command line when you launch the script. The launch syntax is
powershell.exe Path\kb.ps1
InputFilePath kbxxxxxx
where Path is the folder in which kb.ps1 is stored, InputFilePath is the pathname of the input file, and kbxxxxxx is the ID of patch you want to search for. Alternatively, if kb.ps1 and pclist.txt are in the same folder in the default PowerShell directory (e.g., D:\PowerShellscripts), you can type
Path>powershell .\kb.ps1
.\pclist.txt kbxxxxxx
where Path is the folder in which kb.ps1 and pclist.txt are stored and kbxxxxxx is the ID of patch you want to search for.
As Listing 1 shows, kb.ps1 starts by executing two commands that you wouldn’t typically see at the beginning of PowerShell scripts. The first command
$erroractionpreference = `
“SilentlyContinue”
completely suppresses error output. By default, when an error occurs, PowerShell issues an error message, then continues to the next line. When you set the $Error- ActionPreference automatic variable to SilentlyContinue, the processing continues but an error message isn’t issued. Suppressing error messages eliminates distractions for the administrators when they’re reviewing the onscreen patch-status report. Because kb.ps1 is a tried-and-true script that we’ve been using continually for the past 6 months, the benefits of suppressing error messages outweigh the risks.
The second command
clear-host
clears the PowerShell window. Typically, the Clear-Host function is used at the end of scripts, but I used it at the beginning of kb.ps1 to clear the screen before any processing begins. Once again, that helps generate a clean, easily readable electronic report for administrators to review.
After clearing the PowerShell window, kb.ps1 counts the number of commandline arguments. If there aren’t exactly two, it displays the syntax for the launch command. If two arguments are present, the script retrieves them, assigning the input file pathname to the $filename variable and the patch ID to the $kb variable.
Using the Get-Content cmdlet, kb.ps1 reads in the names of the computers in $filename, one at a time. For each computer, the script uses Windows Management Instrumentation’s (WMI’s) Win32_PingStatus class to ping the computer. The Get-WmiObject cmdlet with the -query parameter is used to execute the WMI Query Language (WQL) statement that pings the machine. The script determines whether the ping succeeded (i.e., returned a value of 1) by checking the value in the StatusCode property of the Win32_PingStatus class.
If the ping didn’t succeed, kb.ps1 uses the Write-Host cmdlet to log the computer’s name and the message Ping failed. I didn’t use the Write-Error cmdlet to write the ping-failure information because it mangles the information almost to the point of being unreadable. After writing the error message, the script ends so that the Help desk can determine why the machine is offline and fix the problem.
If the ping succeeded, the script uses the Get-WmiObject cmdlet with WMI’s Win32_ QuickFixEngineering class to retrieve the patches installed on that computer. The script pipes the results to the Where-Object cmdlet, which filters the results for information about the specified patch. The results of that filter operation are then piped to the Select-Object cmdlet, which retrieves the patch’s HotFixID and Description properties.
As callout A shows, kb.ps1 checks to see whether the HotFixID property’s value is the same as the $kb variable’s value. If they match, the script writes the computer’s name and the patch’s description in green text. If they don’t match, the script writes the computer’s name and the message Patch not found in red text.
To write the patch-status information, the script again uses Write-Host. This cmdlet writes information directly to the host interface, which makes the output unusable for pipelining. However, we don’t need the output piped anywhere. Equally important, if the script were to let standard output handle the patch display, you’d get lots of pages containing extraneous information. For our purposes (i.e., generating a clean, easily readable electronic report), using Write-Host works best.
As you can see, there’s nothing fancy about kb.ps1. However, it’s shorter and faster than KB.vbs. Plus, the color-coded results make the report easier to read and more presentable for administrators.
