@kratkale You can do the same thing with “Hard Drive” paramater of the Host in question.

If you know the wwn, serial, or block size of the disk you intend, it should figure out which drive you intend. Of course this is specific to the machine you’re attempting ot image, but this same functionality is possible in FOG.

I have the same HP Z2 G1i – it works for me with
sanboot.PNG
“SANBOOT”
The BIOS also seems very finicky to me—I even disabled all NVMe devices and Windows and enabled only network boot—but Windows resets the setting. Setting a BIOS password didn’t change anything.

@hammerc807

This is what I had to do, was in the same boat.

Go to /opt/fog/.fogsettings

Remove the sysv-rc-conf entry from the packages line from there. (Reference: https://github.com/FOGProject/fogproject/issues/797)

Lower down in the same file, set “php_ver=‘8.2’” to “php_ver=‘8.4’”

Exit and save.

Then I had to remove symlinks in Apache that were using the old php modules:

sudo unlink /etc/apache2/mods-enabled/php8.2.conf
sudo unlink /etc/apache2/mods-enabled/php8.2.load

Then add the new links:

sudo ln -s /etc/apache2/mods-available/php8.4.conf /etc/apache2/mods-enabled/php8.4.conf
sudo ln -s /etc/apache2/mods-available/php8.4.load /etc/apache2/mods-enabled/php8.4.load

@bond007fink A new version of ntfs-3g, which is the parent package for ntfsresize, has finally been released last week. That’s after 4 years of no updates. If all goes well, I will be doing an experimental release this weekend with the newer package.

As @Tom-Elliott stated, this is not an issue with FOG itself, instead FOG relies on tools like ntfs-3g and partclone to handle all the NTFS filesystem changes. Hopefully this updated version fixes your issue.
I will update the thread once the experimental release is available.

@Tom-Elliott

This is fixed, too, and can be marked as solved. The reason was stale WOL power action. I deleted those manually from the DB, and the CPU went back to normal.

I’m not sure about the SSH and replication issues, but I deleted my physical storage node, and I’m not seeing those anymore.

@mashina Ah, thank you for that and sorry for my denseness. Indeed this is a bug and I’ve replicated and figured out where this is occurring.

I have pushed a few mroe adjustments in an attempt find and fix more bugs (thanks AI review and whatnot)

Please pull and let me know if there’s any other issues.

@mashina
I’m still having huge problems. Updated Fog. Rebuilt the Windows 11 image. … Windows 11 changes the boot order.
The image works for a few days—then it stops working.
Multicast only works sometimes—just now the image worked with one computer, but not with three in multicast—it just stops at 94% for no reason. What surprises me the most is that with this computer, only BOOT EXIT TYPE: SANBOOT works. I don’t understand much about this, but it seems wrong to me.

@Bhav There is not at this point in time.

Secure boot is meant to be… well, secure.

There are people attempting to get there, but I can say, right now, FOG just isn’t there right now.

On the forums you can see this at:
https://forums.fogproject.org/topic/17949/fog-secure-boot-with-shim

Which might help lead you in the direction as well hopefully.

@toalalife

@KMEH Well, it must be another image I couldn’t attach.

Which image are you referring to here? bzImage? If so bzImage is the filename of the kernel.

I’ve honestly dedicated many hours to this without success. I’ve seen another project called foguefi (https://github.com/abotzung/foguefi), compatible with Secure Boot directly, without using MOK, and it works. It’s just that when you select options like quickreg or deploy an image, it takes forever. But the point is, it works with recursive boot enabled. It shouldn’t be that complicated.

Yeah, I am aware of foguefi (I actually reference it in the original post), but I think you are slgithly mistaken, it actually is also using MOK. It chains grub via shim (which is similar to what we’re doing here, chaining iPXE via shim) and then re-implements the iPXE menus into grub. This works great, but requires a lot of maintenance effort if anything changes upstream in FOG.

As you can see that project is currently archived, but FOG is also reasonably stable, so I’m not sure how long it will take for a breaking change to stop it from working. Part of the reason I devised this method is that (although hackier to implement on a user level), it doesn’t rely on any functionality that isn’t already present in standard FOG server itself for the most part and it could reasonably be implemented by the FOG developers should the wish to add it.

Anyway, do you have a list of commands to configure it? Maybe I or the AI ​​is missing something.

Sorry I’m a little lost here, commands to configure what? foguefi? If so I’m not sure, I haven’t used it myself so you’d have to ask the developer. If it’s commands to configure my method, the guide above should have everything you need.

@heix75 No.

Not because it’s not a QOL improvement (it is) but because what you see in 1.5.x.x is NOT what is the end goal.

If you want to see what the UX is moving toward install the version of FOG on working-1.6.

You’ll understand (I hope), then, why I’m saying this.

@pillbox1234567 Thanks for the suggestion. I turned Viridian off on the cloned VM, and it did not make a difference.

Interesting note that you’ve seen it cause boot issues on certain Linux VMs. I have not experienced that behavior yet but will keep an eye out for it.

@Preston Wow that was a super answer, thanks Preston! I’ll give this a go in the week. 🙂

@Exsival When you say “Causing the fog server to crash” what exactly do you mean?

If it’s just a “saying” I can understand that but from a technical standpoint what this means is completely different and would need logs. If all you mean is the machine in quesiton is hanging, that’s a different issue altogether. I highly doubt the whole FOG Server is crashing.

@Infojoe I forgot an important information : version 1.5.10.1826

@lucamathuse No, I would not move the files manually yet.

If the capture reports “Cloned successfully” but the task remains In-Progress, then the capture itself may have finished, but the post-capture process is not completing correctly. Under normal conditions, the host should finish the task and then reboot or shut down automatically based on your settings.

So, before moving anything by hand, I would want to know whether the host is actually stuck there indefinitely or whether it is still finishing something in the background. If it never updates the task status, that could also explain why your image set ends up incomplete or inconsistent.