@sebastian-roth Thanks Sebasian I appreciate the input and had a feeling you would say that. Was worth the ask though! Hats off to ya! Cheers

@rogercrf Ok great (well not really, but now we know where to look). The PXE booting bits are all fine.

The issue here is that its possible that FOS Linux doesn’t have the driver for this network adapter.

What I want you to do is cancel this deployment task in the FOG web ui.

Lets first make sure the kernel actually got updated like we need.

from the fog server linux command prompt key in the following
file /var/www/html/fog/service/ipxe/bzImage (case is important). Hopefully it reports the kernel version is 5.15.x (where x is something larger than 50). If yes then we need to debug this kernel driver issue.

To debug this issue manually register this computer with the FOG server web ui. You will need to know the mac address of the computer. Once you manually register this computer with the FOG server, schedule a caputer/deploy task. Doesn’t matter, but before you hit the deploy task button tick the debug checkbox. Now schedule the task and pxe boot the target computer. It should immediately jump to imaging and not present the iPXE menu. You will still see the ip error message but it will drop you to the FOS Linux command prompt. Key in the following commands.
uname -a
ip a s
lspci -nn | grep -i net
grep -i firmware /var/log/messages

Take a clear picture of the output and we can debug more. But lets hope the kernel didn’t get updated and just an update will fix it.

@taylorcockrell Well there is no easy answer here. If you need secure boot enabled in your environment then you can create a self signed key and apply it to each workstation. Then you can sign both ipxe.efi and bzImage with the same key. Once that is done you can secure boot using FOG. I created a tutorial on this on the steps needed. For an opensource project its a bit impractical to get microsoft signed kernels and efi boot loaders to do it any other way. I wish there was a better solution.

In the case of the hardware, I know for Dell hardware you can use a Dell offered utility to modify the firmware from within the host OS. Thinking that you can turn off secure boot (which will break bitlocker, but you will reimage the computer anyway) then reboot the computer into PXE booting with FOG.

If you require an IT tech to sit in front of the computer to image it, then they can simply turn off secure boot and then boot into pxe booting via the uefi boot manager. The imaging tech would have the access and capabilities to disable secure boot prior to imaging.

@castelon The quick answers are yes, yes and no.

Ideally you will have the answer file on the windows golden image before you run the sysprep command because when calling sysprep you will point to the exact location of the answer file. Then shut down and capture with FOG.

While I was developing my golden image I wrote a post download script to open the windows disk and replace the answer file. I did it this way to avoid having to restore, update teh answer file then sysprep and recapture with fog.

While yes you can deploy it with a snapin, it won’t do you any good because winsetup needs that answer file BEFORE snapins would run. Its a chicken and egg situation.

Hi,
I force computers (with UEFI) to boot from HDD (Windows bootloader) with this command in snapin:

#Booot from HDD
#Snapin works only with Powershell x64 Snapin Template?
#Use at your own risk! No warranty!
c:\windows\system32\bcdedit.exe /set “{fwbootmgr}” displayorder “{bootmgr}”

We use it after imaging because PXE boot is too slow.
Peter

@multipass ok would you follow this tutorial to capture a pcap of the pxe booting process. https://forums.fogproject.org/topic/9673/when-dhcp-pxe-booting-process-goes-bad-and-you-have-no-clue

Upload that pcap to a file share site and DM me the link using FOG chat. I’ll take a look at it for the next steps. The capture filter will only collect the dhcp process and no other traffic on your network.

@rtarr said in Boot Loop after Sysprepping a Golden Image VM:

It then boots to the login page for the local admin where install packages are located on the desktop.

I’m glad you have it worked out.

To extend what you have done just a bit more (if needed). Look at my unattend.xml script there are sections where to add local keyboard support (if you are not in the US) as well as it should have autoadmin login with a login count number that will cancel the auto admin login after X number of logins. If not you can add it to your unattend.xml file. That in conjunction with the run firstlogincommands where you can install any applications in an unattended manner as the local admin. The last firstlogincommand should be shutdown -r -t 0 that will reboot the computer leaving it at the login window.

I just looked and my posted unattend.xml file in the link does not included this section. I must have removed that before posting. But above is enough info for you to google it.

Thank you @george1421 .
It was really perfect.

Since I have some issues with Sysprep, I changed my way. So let’s forget the previous story for imaging.

What is my scenario?
My clients (coffeeshop) every day login to the systems. I need to reimage machines every time they log out from systems. So that they can get fresh installed windows after turning machine on.

So what I did do for machines?
I installed windows 11 on my workstation which is Dell 3420 Precision Tower. Then, I installed a few apps like chrome, pdf reader, vlc. Finally, I put this script to the windows task scheduler in order to execute it right after the computer boot up or shut down. this script helps me to keep network boot as the first boot order (Since windows manipulate the first option to itself while it should be pxe because of the next boot). This script will run on every boot and every shutdown(just making sure pxe is first order).

What is my issue?
I simply captured the image without anything else llike sysprep. I deployed this image on the original machine which I installed windows 11 from scratch. Now Windows 11 can not boot up. First, it goes to a blue screen and it says “BAD CONFIG SYSTEM INFO”, and then it goes to the automatic recovery page and SrtTrail.txt says: “Check for installed LCU Error code: 0x3f1”

It seems fog can not capture windows 11 images correctly or can not restore the same image on the same machine.

If fog is fine, what is my mistake? :))

@kallum-redgrave said in windows 10 21H2 FOGService not starting:

sc config FOGService start= delayed-auto
shutdown -t 0 -r

Just for reference: https://docs.microsoft.com/en-us/windows-hardware/manufacture/desktop/add-a-custom-script-to-windows-setup?view=windows-11

*Warning

You can’t reboot the system and resume running SetupComplete.cmd. You should not reboot the system by adding a command such as shutdown -r. This will put the system in a bad state.*

Since I don’t use the fog client on my campus I can’t verify. The original post can be read a few ways.

Is the SC program failing to configure the FOG Service to start? The SC program is working correctly and the service is set to start but the FOG Service fails to start?

Which is the case?

@jdakel did u try CCTK by Dell ?
it allows you to modify bios from your chair

@dpullin said in FOG service invalid host:

The issue seems to be when the name in FOG UI doesn’t match the hostname of the machine (because I changed it in the UI after the fact) - this has started to crop up on other machines due to user error (techs giving machines the wrong names before sending them out the door) - then we try to fix it by changing the name in the UI, then it gives the “invalid host” error as described above.

I can’t seem to replicate the issue as described using the following steps - please verify if you do things a different way:

Check fog.log to see communication works fine Make sure AD credentials are set and correct for the test host in the FOG web UI Change Host Name in FOG web UI Wait for machine to reboot Check hostname (Windows GUI as well as CMD -> hostname) and fog.log (see below). ------------------------------------------------------------------------------ --------------------------------HostnameChanger------------------------------- ------------------------------------------------------------------------------ 4/16/2022 7:06:09 AM Client-Info Client Version: 0.12.0 4/16/2022 7:06:09 AM Client-Info Client OS: Windows 4/16/2022 7:06:09 AM Client-Info Server Version: 1.5.9.139 4/16/2022 7:06:09 AM Middleware::Response Success 4/16/2022 7:06:09 AM HostnameChanger Checking Hostname 4/16/2022 7:06:09 AM HostnameChanger Renaming host to Newhostname 4/16/2022 7:06:09 AM HostnameChanger Renaming host inside existing domain binding 4/16/2022 7:06:09 AM HostnameChanger Success, code = 0 4/16/2022 7:06:09 AM Power Creating shutdown command in 60 seconds 4/16/2022 7:06:09 AM Bus Emmiting message on channel: Power ------------------------------------------------------------------------------

Works fine in my test setup. Do you use plugins or other special things in your FOG setup? Other than that I con only imagine is has to do with the wireless MAC you seem to be using. Is this the primary MAC?

That work now, thank you !

hello

we have dell 3540 using UEFI boot ( legacy boot disabled everywhere)

secure boot has to be disabled dhcp is made via pfsense with ipxe.efi master has been made with UEFI ( fog service has to be stopped before sysprep , otherwise bootloop after deploy / don’t know why) STP disabled on every switch

Works fine here (single or multicast)

@george1421 yep! That’s how I have it set up. Works beautifully now

re: bitlocker - yeah, that makes sense regarding mismatched TPM signatures. Luckily I don’t have to worry about that at the moment 🙂

@mmaridev Do you use @JJ-Fullmer’s powershell FogAPI module by any chance? I remember digging through an issue with him that sounds kind of similar and we got to the point where it turned out to be caused by a powershell script he used. Can’t remember all the details anymore.

@dashwell Please use the latest FOS init as posted in your other topic to see if this is still happening.

If it does I would like to see the contents of d1.partitions and d1.minimum.partitions from that image.

@george1421 Our current solution was to isolate the two(technically 3, since the fog server is running in an Ubuntu Virtualbox) to a completely separate domain.

I brought in a spare personal router, hooked the two physical machines up and set the DHCP settings on that. The DHCP problems are no more, for now at least, and whenever necessary, or available, I can always re-bring these two back online after the image capture is done.

Thank’s a lot for for quick response. We’ll test and comme back to you.
Have a good days
Francis