Powershell move-item does not remove some 3D png files from source

Question

I have a Powershell script issuing the Move-Item command to move some 3D CAD directories and subdirectories from one network share to another. About 300 parent folders a day. I am having the most strange problem. All parent folders are copying to the destination just fine, but about 10 of the 300 objects are not being removed from the source. It seems that any subdirectory that contains a type of png scan file, will not remove from the source; it copies fine, just doesn't remove from source. No error is reported in the -verbose log output. The file sizes are not large, < 50MB each. Doing some testing, I also noticed that if I run the script once, the problem occurs, but if I run the script again with no changes, it moves the remaining objects that it did not move on the initial run. Another observation is that if I run the script where the source and destination are on the same share, the problem does not occur. The problem only occurs when source and destination are on different shares. Same problem using both versions 5.1 and 4.0.

I also tried using some simple png files I created myself just by saving a jpg screen print as a png, and those png files copy and remove fine. So something is special about these 3D CAD png files. I even tried copying off the parent folder to share to make sure no one or no program is locking the original files, and the problem still happens.

I have been successful at replacing the move-item command with a robocopy command, and it works perfectly, so I could use robocopy to solve my problem, but I want to figure out why the move-item command is having this problem. My only theory is that if on the same share, no data is actually being moved, only the reference to the file, ie pointer, ie FAT (file allocation table) is being changed. When crossing shares the actual file data has to be moved, so something is funny about these files that prevents that from happening. But as stated above, if I run the script a 2nd time, the remaining files copy and remove fine. So that leaves a little unexplained.

Just wondering if anyone has any ideas.

See the code samples below, and also some screen prints of the source folder structure before and after.

$sLogPath = 'C:\PowerShell_scripts\test\robocopy_test.log'

$StagingFolder = '\\3ipbg-fs-p01\PSP_Inbound\Test\3Shape_Auto_Create\Suspense\test_case_01'

$FinalFolder_M = '\\3ipbg-fs-p01\patient specific mfg\3Shape_AutoCreate\Auto_Job_Create_Test\Inbox_Manual_Jobs'
$FinalFolder_M_robo = '\\3ipbg-fs-p01\patient specific mfg\3Shape_AutoCreate\Auto_Job_Create_Test\Inbox_Manual_Jobs\test_case_01'

Move-Item -LiteralPath $StagingFolder $FinalFolder_M -Verbose -Force

robocopy $StagingFolder $FinalFolder_M_robo /R:3 /W:3 /E /B /J /MIR /MOVE /LOG+:$sLogPath

enter image description here

Answer 1

You are using /R:3 parameter for robocopy, meaning it retries to move the folder up to 3 times if an error is present. I am not aware of such functionality in PowerShell, but you could easily write a small utility function for this.

Assumptions:

1) In the loop code, it checks if the source folder exists, if Yes, it continues up to $retries times with Sleep of 2 seconds.

Function Custom-Move{
    param( [string]$source, [string]$destination, [int]$retries)

    $i=0
    while($i -ne $retries){
     Move-Item -LiteralPath $source -Destination $destination -Verbose -Force -ErrorAction SilentlyContinue

     #test if the folder exist, if not meaning that the folder was moved.
     if(-not (Test-Path $source)) {break} 

     $i++
     Start-Sleep 2
    }
}

Usage:

Custom-Move $StagingFolder $FinalFolder_M 5