Misbehaving Get-ChildItem Operation in PowerShell


Misbehaving Get-ChildItem Operation in PowerShell

There is a back-end SQL DB contains “managed folders” in the form of UNC paths. Using SQL queries in PowerShell I have a loop that will work it’s way through these folders and run a GCI operation against them to work out how much disk space they are using.
$managedFolder = “\\server\share\folder\subfolder”

For the sake of the question, $managedFolder is declared as above. The failing command below:
$diskTrendsInitialUsage = “{0:N2}” -f ((Get-ChildItem $managedFolder -Recurse -Force | Measure-Object -Property Length -Sum).Sum / 1GB)

Now if I were to run this command manually in PS console it’s fine, it pulls back data. But as soon as it’s packaged in a script, it fails with the below error. The folder is accessible from the server, as it works fine from a local PS console session.
ERROR: Get-ChildItem : Invalid Path: ‘\\server\share\folder\subfolder’.
AddManagedFolder.psf (17): ERROR: At Line: 17 char: 42
ERROR: + $diskTrendsInitialUsage = “{0:N2}” -f ((Get-ChildItem $managedFolder -Recurse …
ERROR: + ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
ERROR: + CategoryInfo : NotSpecified: (:) [Get-ChildItem], ArgumentException
ERROR: + FullyQualifiedErrorId : System.ArgumentException,Microsoft.PowerShell.Commands.GetChildItemCommand

I’m stumped.


Solution 1:

The problem with your path is that it does not have any indication about which provider to use, so PowerShell just use current one. And if current provider is not a file system provider, then it will fail. So you need to specify provider in path, to allow PowerShell to choose right one regardless of current provider:

$managedFolder = "filesystem::\\server\share\folder\subfolder"

Solution 2:

My guess is your are using the SQL PS cmdlets prior to running GCI, this is changing your provider path to SQL: which is what is causing GCI to be unhappy.

Prior to running GCI do cd c:\ to change the path back to the file system and GCI will work.