I was not able to access a newly joined Windows 7 computer on our domain from our Windows 2000 Server.  I just received that error.  “Cannot perform the requested operation”  Sharing was setup ok, permissions and security set ok, firewall was turned off.  I saw an Event ID 2017 on the Windows 7 system, too. Something about “unable to allocate from the system nonpaged pool”.

I found this online. Add the MaxNonpagedMemoryUsage value below in the client’s registry. (Windows 7 system)  Make a restore point first so you can undo if any problems arise.

HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\LanmanServer\Parameters\MaxNonpagedMemoryUsage

DWORD VALUE: 0xFFFFFFFF

Next, restart the “Server” service on the Windows 7 machine.  Then go back to Windows 2000 Server and try to connect to \\window7machine and see if it can access it now.

This worked great for me, hopefully it works for you too.  However, it may be wise to increase that value in small increments from the default, which I THINK is 0×100000.   I just maxed it out and on my client system to test, and it’s working fine.  Just thought I’d make note of it.  If you have further problems, or this doesn’t fix the issue, remove the value and reboot.  This should return it back to the default.

Not too often would you need this, but I just ran into a situation where it was handy to get AD off an existing, but inactive, domain controller.  I couldn’t get dcpromo to uninstall the services and demote the darn thing, because this server thought it was still active in the domain.  I had physically removed this server and manually cleared it from the active domain.  Why would you do this, well, that’s a long story, but I wanted to, let’s just say that.

Here’s how you demote a DC with dcpromo, forcefully.

“dcpromo /forceremoval”

I did this on a 2008 Server, worked great!  All these years and I hadn’t needed to do this, and didn’t even know it could be done.  Always something new to learn, huh!

I’ve moved a few virtual machines off of both ESXi and VMWare Server recently. (was testing VMWare, which I do like.) I moved them into VirtualBox while I test OpenSolaris, which I am REALLY liking… but off topic. Anyway, once I got the vmdk hard drive files copied, I created a new machine config in VirtualBox and added the vmdk to VirtualBox. (which is very cool I must say, you can use vmdk’s or vhd’s in VirtualBox.) Once the OS booted, in both XP Pro and Win2008, the performance was terrible and had no network. I realized that the VMWare Tools were probably still installed. Should have removed them first, but whatever. When to control panel and tied to uninstall, but I got some error that “No VmWare products were detected”, great! That’s nice, you can’t even uninstall after they are moved off the VMWare host services.

I added the VMWare Tools ISO to that VM and opened a command prompt. Then I found the “setup.exe” or “setup64.exe” files on the cd. Run those in the cmd window with a /c switch. (example:  ”setup64.exe /c” )  This will remove the Tools.  I installed the VirtualBox additions and then reboot.  On my Windows 2008 server it crashed on first reboot, then then I tried again and it loaded up just fine.  Network works and everything! Yay!

EDIT 7/23/09 :

I noticed on one Win2k system that the VMWare uninstall didn’t fully remove the networking drivers.  When I rebooted, the system tried to install the AMD PCNet adapter again, only it said it was “VMWare  …”   What I did to fix that…  I shutdown the Guest OS, went to settings and under Networking I selected the Intel adapter instead of the AMD.  This worked, but also required that I download the driver from Intel and install it manually on the Guest OS.  Everything worked fine though, and I had Gigabit networking!

I’d still like to know how to get rid of ALL the VMWare stuff though.  Probaby just need to find some INF’s for something in Windows.

This drove me nuts!  Why Microsoft had to take something totally simple in Windows 2000 and make it a complicated thing is NOT beyond me!  This is MS we’re talking about!  Of course it’s not easy with newer versions.

Took me a little bit, but here’s the commands I used on our primary domain controller, and it’s working great. that first w32tm command is all one line.

w32tm /config /manualpeerlist:”0.pool.ntp.org 1.pool.ntp.org 2.pool.ntp.org pool.ntp.org”,0×8 /syncfromflags:MANUAL /reliable:yes

w32tm /config /update

net stop w32time

net start w32time

w32tm /resync /rediscover

That should do it. However, always make sure you firewall is open to port 123 outbound!  I initially was receiving this error after running a “w32tm /resync” :

The computer did not resync because no time data was available.

In my case, that was caused by my firewall blocking port 123 for NTP traffic.  Go figure, we’ve been running this particular network for probably 2 years with that firewall blocking port 123, and only now did someone come and ask “why is our computer time off by 6 or  7 minutes?”  This is when you say, “Welcome to the world, can I help you?”  (Good old Beavis)  Well, at least we got our server configured better as a “reliable” time source with the right ntp.org pools.