@whity said in Hostname while joining to the domain:

I am planning to have FOG client installed in the golden image. Is it mandatory to have the Sysprep done before capturing the image?

Sysprep is a standard/typical step needed to make the reference image ready to install on multiple hardware. There are too schools of thought there. If you are using the idea that you will create one golden image per hardware model then you technically don’t “need” to run sysprep. But if you want to have one golden image for all hardware platform then you need to use sysprep to neutralize the golden image so that it can install on multiple hardware configurations. I can say I always run sysprep with my golden images. The decision to run sysprep or not is not related to FOG. FOG doesn’t care if the image has been syspreped or not.

Also, my understanding is that the FOG client is needed in the golden image to aid the domain joining while deploying the image? Is that right?

The fog client is needed if you want FOG to connect the target computer to AD, name the computer as defined in FOG, as well as if you want to run snapins or other fog managed services. In my case I don’t use FOG to connect the target computer to AD because I have a complex domain structure. I do use FOG’s post install scripts to calculate the proper AD location for the computer at deployment time and then update the unattend.xml file with the proper AD location and calculated computer name, then when Windows OOBE runs it looks at the unattend.xml file and connects the target computer to AD in the proper OU location. This is having windows setup connect the computer to AD and not FOG.

I think I hunted down the culprit. I installed each piece of software one at a time and it looks like it is the new cloud version of Sophos that is giving me the error. Removed it from the image and everything seems to have went through without an issue.

@Sebastian-Roth

Where is the setting to set hostname even if a user is logged in?

I was looking for it and stumbled upon my AD settings which is what I ended up fixing instead that fixed the issue.

@Quazz said in Can't deploy latest win10x64uefi with 1.5.5:

Wow, you’re so fast with answers! I really appreciate that!
As I’ve said in the firs post, I’m using Hyper-v second generation VM based on UEFI.
AAAAND! It seems that I’ve got some progress(really hope so).
What i’ve done:
Downloaded latest binary zip of rEFInd: https://sourceforge.net/projects/refind/files/0.11.4/refind-bin-0.11.4.zip/download
extracted /refind/refind_x64.efi ; drivers_x64 from this folder
Renamed refind_x64.efi to refind.efi, renamed drivers_x64/ntfs_x64.efi to drivers/ntfs.efi, renamed /drivers_x64 to /drivers.
Chmod 644 chown fog:www-data to this files|folders, moved them to /var/www/fog/service/ipxe, removed comment symbol from this line:
#scan_driver_dirs EFI/tools/drivers,drivers
in /var/www/fog/service/ipxe/refind.conf.
And I’ve just deployed my image with only one error at the end:
“failed initialization of the interactive login process” or something like this, it’s translated from Russian.

i’d split this up - use GPO or at minimum registry to achieve this then snapin to push out relevant xml start menu layout

Registry:
[HKEY_LOCAL_MACHINE\SOFTWARE\WOW6432Node\Policies\Microsoft\Windows\Explorer] - 64bit location
“LockedStartLayout”=dword:00000001
“StartLayoutFile”="C:\PATH\TO\START\MENU\LAYOUT\.xml

you can partially lockdown start menu aswell so you give users a standard put they can pin their own stuff too (not allowed to edit what you’ve set)

the snapin could just handle which start menu they get

i.e. Admin-> this startmenu.xml
IT-> this startmenu.xml

just when they reach the location set in registry name it the same so like:
IT-Startmenu.xml copies locally to startmenu.xml
Admin-Startmenu.xml copies locally to startmenu.xml

Hope this makes sense

@cristianvalente said in ADK provisioning package settings won't apply after an PXE Boot:

Is somebody familiar with ADK and PXE?

Yes and they are not related. Provisioning packages are deployed from inside the OS. FOG doesn’t now or care about windows applications. Are you trying to deploy a provisioning package as a fog snapin?

Hello!
Yes, my idea is fog installed widnows can keep the user’s other partitions, because I want to keep the user’s other disk character data

@jflippen Thanks. I will give that a try and let you know. I really appreciate the responses.

@Les We need more information to be able to help. The partclone.log file only exists on the client. So you might schedule another capture task but tick the checkbox for debug task this time. Start the client and when you get to the shell start the capture via command fog and step through till you hit the error. Make sure you keep your eyes open for errors just before the “Partclone failed” message! Now after the error happens you should be send back to the shell. See if you can get the error message from the log file.

As well you might post more information about your FOG version, the image settings (resizable or not, partclone/partimage, OS type). Plus you should find the partition tables captured in a folder /images/dev/XXX (XXX should be the MAC address of your client) on your FOG server. Run cat /images/dev/XXX/d1.partitions /images/dev/XXX/d1.minimal.partitions and post output here.

@Sebastian-Roth I have tried to run it as a domain log-on script but it didnt work, I can still install some apps ( prepared using SFX maker as silent install through Fog Snapin) but none of the script is working. I would appreciate if you kindly advise me? Many thanks

@kickers56 Please try to set the domain password using Incognito mode in your browser, see if the issue persists. Browsers will auto-fill and auto-complete password fields as a feature, and this has on many occasions been the culprit in strange password issues with fog.

@x23piracy Thanks, that’s good to know. Although I guess it might be that $ special case causing the issue for the other users as it is used for variables in PHP, as line end in regular expressions such things.

@David-Raymond From what you describe this is definitely not a FOG issue. We might point you the right way to solve this but I am sure there are other forums out there knowing a lot more about that kind of stuff.

Maybe start reading here: https://social.technet.microsoft.com/Forums/en-US/fe8901f2-f8d9-44aa-8eb4-b6b853ada623/quotrpc-server-unavailablequot-while-attempting-to-join-domain?forum=winserverDS

Sounds like a DNS issue. You might want to run the mentioned DCDIAG command on your Windows AD server to see if it is all setup correctly.

@iekozz There are two commands you need to use when removing applications. The first one removes it from the current user the second one removes them from the system. You do this as the administrator before any one else logs into the computer. After the script is run, the next users that login will not have the application on the desktop or start menu. That is the way at least it use to work. I have not done image design in a few months. MS could have changed things (yet again) with 1809 that I’m not aware of.

If anyone is interested this worked for me. I found this in a MS thread. I placed this in my Windows\Setup\Script\SetupComplete file just above the fog service enabled command.

del /F /Q C:\Users\Default\AppData\Local\Microsoft\Windows\WebCacheLock.dat
del /F /Q C:\Users\Default\AppData\Local\Microsoft\Windows\WebCache*.*
del /F /Q C:\Users\Default\AppData\Local\Microsoft\Windows\INetCache*.*

There are 2 ways that I go about this.

Put the commands you need to execute in the setupcomplete.cmd batch file (like creating accounts and such). Just make sure you delete this batch file after it runs to ensure you don’t have any password leakage.

We use the first run section of the unattend.xml file for actions that need to happen inside a windows profile. We setup the unattended login function with a count of 1. The second to last step of the first run section changes the login screen to our standard company branding (so we know the imaging process is complete) by updating a few registry settings, and the very last step of the first run section is to issue a shutdown.exe -r -t 15 -c "This system will reboot in 15 seconds" command. This forces the computer to reboot after all of the other actions are done.

@Scott-Boucher Thoughts? Yes… switch back to win7.

The only way I know to make this work is to customize a start menu design to your liking. Then export it to its xml file format. Then copy that xml file to your golden image. The final bit is to create a GPO policy to force the computers to load its start menu from the exported file you created before.

You may be able to do this with the microsoft provisioning package, but I haven’t done that before.

A quick google search on “win10 customize start menu” should point you into a direction.

@Joe-Gill said in Keeping WIN10 Images Updated (Windows Updates):

Thanks for that info. I’m looking into setting up a server with MDT on it. Hopefully that will help these issues.

FWIW: For MDT you don’t need a server class machine or OS. You can use desktop os and computer for this task. Virtualize if you can, but lower class physical hardware for “MDT server” is usable.

Well after a bunch of research and testing, this little tool works pretty well.

http://www.forensit.com/support-downloads.html

Defprof

A little utility that you setup one profile exactly how you want it and then this will copy all those profile settings to the newly added users. Works for domain users as well.

Alright. Thank you everyone for your time and answers.