Glenn Sullivan's questions

Does anyone know how to change the loadbalanceinfo string for a RDS Session Host Remote Desktop collection?

Use case: we had a case today where something got corrupted in our RDS collection, and there was little info available on the errors that were being thrown except "delete the collection and recreate it." Given that we had 70 people unable to work, that was our best option.

That is what I did, but, of course, it had a different loadbalanceinfo string in the RDP file for the new collection. If we were using RDWeb or RemoteApp deployment to allow access, that's no big deal, but we have a large amount of HP ThinPro Linux thin clients, so it meant changing a test client's connection properties to reflect the new loadbalanceinfo string and then deploying new connections to all of the thin clients so they could connect to the new collection.

Additionally, we had many users who had downloaded the RDP file from RDWeb for the old collection to their desktop or something, and all of those had to be replaced as well.

In the case that this happens again, is there a way to change that string on the new collection so that it matches the old collection's string and no connection replacement is necessary?

I am migrating my users from a Server 2012 R2 farm using Roaming Profiles to a Server 2019 RDS deployment using User Profile Disks. I have created a script that will copy the "important" folders from the UNC where the roaming profile is stored into the user's new UPD, and have deployed that script via GPO as an immediate schedule task that only is created if a particular "Profile migrated" flag is non-existent (which the script obviously creates).

It is works very well under "normal" circumstances.

But, if a user happens to log on and their UPD is locked (an admin user primarily, who has the rights to log into multiple session host servers at once, and knows that this might happen under those conditions) and they get assigned a temporary profile/UPD, I want to detect that and NOT run the script.

I have not found a reliable way that I can check to see if the current session is using a temporary profile that I can use in "Item Level Targeting" to further filter and prevent this scheduled task from being created if the user is running under said condition.

Any suggestions? Registry settings seem to be the most likely candidate, but I haven't found one that works yet.

So I have an old Nexus 3524 that is set up with a bunch of port-channel pairs connecting to three different ESXi hosts and four different NetApp nodes. I am replacing that with a pair of Nexus 9000's in a vPC configuration.

I've created a new port-channel to flow traffic between the old Nexus and the new stack and it's working fine for management traffic. I've also created new vPC port-channels on the new pair of switches to host each of the existing pairs connected to the old switch.

Is there a good non-disruptive way to move the pairs of ports from the old Nexus port-channels to the new vPC Nexus 9000 port channels? For the NetApp, I can just migrate the lif off of a port group that I want to move, move them, and then migrate it back home. That'll work for NFS and CiFS traffic, but iSCSI traffic and intercluster traffic might not like it.

I guess I'm most curious what will happen if I move one of the ESXi connections at a time to the new Port-Channels. Anyone tried it?

I'm going to do it in a maintenance window, but I'm just looking to hear what folks have experienced and learn what to expect.

I am looking for a way to automate nightly backups of my VMWare ESXi Configs for disaster recovery.

Ideally the script would connect to my vCenter server, poll it for hosts, and then backup the config in a logical directory structure, including the current ESXi version running, since config backups can only be restored to a machine running the exact same build.

Is there such a script available somewhere?

I have a very strange situation. I have a Windows 2012 R2 VM with mounted thin provisioned LUNs located on a NetApp FAS2552 filer (running cDOT 9.4P2), and the server is not sending SCSI UNMAP commands when files are deleted from the LUNs.

Things I have tried:

• C:\Windows\system32>fsutil behavior query disabledeletenotify (returns 0)
• Restarted the client OS
• dismounted and remounted the LUN

I have verified that there are no UNMAP commands being received by the toaster by doing a statistics run on the LUN while deleting files, and looking at the counters as suggested here: https://kb.netapp.com/app/answers/answer_view/a_id/1002617/~/faq---windows-scsi-unmap-

I opened a case with NetApp, and they are claiming that, because the host is not sending unmap commands, it is an issue that I need to take up with Microsoft.

So I came here first... anyone have any idea how to figure out why this host is not sending unmap commands when files are deleted?

I have a public folder that contains "Notes" type items on my Exchange 2016 server. A user inadvertently deleted one of those notes.

Mail type folders have the option to "Recover Deleted items" in Outlook, but Notes type folders do not.

Does anyone know an easy way to just pull this item out of the Deleted Items? I have retention turned on, and we are well within the retention period.

We migrated a SQL SSRS server from one machine to another. In order to make it so that everyone's shortcuts and favorites to specific reports still work, I shut down the old SSRS server and created an IIS site on that server that uses URL rewrite to redirect the request to the new server. It works great.

Now we want to, over time, contact the users who have not replaced their shortcuts and favorites with new ones on the new server and get them to do it so that we can shut this old server off.

I thought it would be as easy as making it so that the only authentication method available was "Windows Authentication" (which is how this specific site was set up to begin with) and then look in the log files. But all of the log file lines have no username... which makes me believe that the URL rewrite is taking place BEFORE the authentication.

Anyone have a workaround that would force the authentication to the old server to work, so I can get usernames in the log files?

I am building a script to remotely install the PSWindowsUpdate PowerShell module, open the correct firewall ports, and then run a command to install waiting updates. In order to install the PSWindowsUpdate module, I need to install an MSI on some of my machines to enable the "Install-Module" commandlet.

If I run the following code manually, in an admin PowerShell session on my local machine, it works fine:

$RemoteMachine = "DCSvrRDS16"
write-host "Server Name is: $RemoteMachine"

#Copy the MSI Local to the computer
Write-Host "Copying MSI locally"
Copy "\\dcsvrstorage2\software\microsoft\powershell\PackageManagement_x64.msi" \\$RemoteMachine\c$\

#Run the MSI remotely
Write-Host "Running MSI"
Invoke-Command -ComputerName $RemoteMachine -Credential $domaincredentials -ScriptBlock { c:\windows\system32\msiexec /i c:\PackageManagement_x64.msi /quiet /lvx* c:\PackageManagement.log } 4>&1

But if I run THIS full script IN THE EXACT SAME ALREADY OPEN PowerShell SESSION, the MSI does not install and the log file is not created:

Import-Module PSWindowsUpdate
$domaincredentials = Get-Credential


Function ProcessMachine($RemoteMachine) {

    write-host "Server Name is: $RemoteMachine"

    #Copy the MSI Local to the computer
    Write-Host "Copying MSI locally"
    Copy "\\dcsvrstorage2\software\microsoft\powershell\PackageManagement_x64.msi" \\$RemoteMachine\c$\

    #Run the MSI remotely
    Write-Host "Running MSI"
    Invoke-Command -ComputerName $RemoteMachine -Credential $domaincredentials -ScriptBlock { c:\windows\system32\msiexec /i c:\PackageManagement_x64.msi /quiet /lvx* c:\PackageManagement.log } 4>&1

    #Install the module
    Invoke-Command -ComputerName $RemoteMachine -Credential $domaincredentials -ScriptBlock { Install-Module PSWindowsUpdate -Force }

    #Turn on the firewall rules
    Invoke-Command -ComputerName $RemoteMachine -Credential $domaincredentials -ScriptBlock { Set-NetFirewallRule -DisplayName "COM+ Network Access (DCOM-In)" -Enabled True }
    Invoke-Command -ComputerName $RemoteMachine -Credential $domaincredentials -ScriptBlock { Set-NetFirewallRule -DisplayName "COM+ Remote Administration (DCOM-In)" -Enabled True }
    Invoke-Command -ComputerName $RemoteMachine -Credential $domaincredentials -ScriptBlock { Remove-NetFirewallRule -DisplayName "Remote WSUS Install" }
    Invoke-Command -ComputerName $RemoteMachine -Credential $domaincredentials -ScriptBlock { New-NetFirewallRule -DisplayName "Remote WSUS Install" -Profile Domain -Enabled True -Direction Inbound -Action Allow -Protocol TCP -LocalPort RPC }

    #Try the update
    Install-WindowsUpdate -ComputerName $RemoteMachine -AcceptAll
}

foreach ($machine in (Get-Content NeedsWSUS.txt)) {

    ProcessMachine $machine

}

It is driving me crazy. I have run ProcessMonitor on the remote machine, and I can see msiexec start and then stop without ever accessing the MSI in the root of the C: drive or trying to open the log file for write.

Has anyone seen anything similar? Client machine is Windows 7 pro, Powershell 4, and the destination machine is Windows 2012 R2 and also PowerShell 4.

Thanks in advance.

I have a single server Exchange 2010 installation.

Three times in the past week Outlook Web Access has stopped working, but ONLY for Internet Explorer clients. The forms based authentication login page loads, but no matter what credentials are entered it always returns the message "The user name or password you entered isn't correct. Try entering it again."

Opening just about any other web browser on the same client computer and trying from there with the same credentials works just fine.

This tells me that there is something wrong with the PREMIUM client, but what?

What I have done in the past that seems to have solved the problem, but only temporarily:

1. Reapplied Update Rollup 4 for SP2 and restarted.
2. Just restarted.
3. Recreated the OWA virtual directory and restarted IIS.

I am reluctant to restart the server now in the middle of the day to try options 1 or 2, but option 3 didn't work this time.

Any suggestions? I have tried with multiple IE versions (7 through 9) with the same results.

I am having an issue where, at seemingly random times, a large portion of my RDP Terminal Server users are being disconnected for no apparent reason. The disconnection is very short lived, and often the RDP client will reconnect automatically, but many of our HP thin clients do not.

I suspect the switch that we have in our data rack, a Netgear GS748T, which is currently serving triple duty, with a portion VLANed for SAN traffic to our Netapp filer, a portion VLANed for vMotion traffic and the rest VLANed for connections to the rest of our standard network from the three vmWare hosts, and CIFS access to the Netapp.

Does anyone know of any non-standard activity logging capability for the GS748T? There is very little in the web interface, and less online.

I believe that this must be caused by some kind of port reset on the ports that feed back to the rest of the network, but without any kind of log, that is difficult to check.

Thanks in advance for any help you can offer.

I have a real strange one... I do IT work for an accounting firm, and because they have such a diverse group of clients, they have many different versions of Quickbooks installed at once. All employees run as "standard Users" (ie, not admin or power users). All PC's are Win7-64 bit, fully patched via a local WSUS server.

Somehow about three weeks ago (maybe a month... perhaps last month's patch tuesday caused it?) all of the users started reporting that the Quickbooks 2011 desktop icon was displaying with a "UAC Shield" on it and they were being prompted for a username and password when they invoked the program.

The strange thing is, the username and password that they were ALREADY LOGGED ON WITH worked to bypass this UAC prompt. But once the program opened no network drives were available, almost as if they had been logged in with a separate copy of their profile but without the logon script running first so no mapped drives are available.

A complete reset of a given user's profile fixes the problem, but that is not a real acceptable solution...

So, any suggestions for ways that I can change the UAC prompting for this application? There has to be something in the profile that causes this to happen that I can look into.

I appreciate any suggestions,

Glenn