In the end I am maintaining a separate image, I was able to get management to let us buy a separate surface go 4 for maintaining the 4k disk image.
I found that bhyve based VMs can be set to 4k blocks but it was cumbersome to get it to boot to fog to capture the image at the end. And when that image was deployed, it did not expand on the surface go.

Some info from the debug session

cat /images/4KDisk-Base-Dev/d1.original.fstypes
/dev/vda3 ntfs

[Fri Jan 26 root@fogclient ~]# cat /images/4KDisk-Base-Dev/d1.partitions
label: gpt
label-id: 9865AAFC-B984-4860-ACF5-4D6F2513747D
device: /dev/vda
unit: sectors
first-lba: 6
last-lba: 16777210
sector-size: 4096

/dev/vda1 : start=         256, size=       76800, type=C12A7328-F81F-11D2-BA4B-00A0C93EC93B, uuid=7C4743B5-7150-4672-B521-7B537528D7E7, name="EFI system partition", attrs="GUID:63"
/dev/vda2 : start=       77056, size=        4096, type=E3C9E316-0B5C-4DB8-817D-F92DF00215AE, uuid=F6216A84-0172-4445-B616-E36DFA20C731, name="Microsoft reserved partition", attrs="GUID:63"
/dev/vda3 : start=       81152, size=    16501760, type=EBD0A0A2-B9E5-4433-87C0-68B6B72699C7, uuid=B6DA06DC-A5D0-434C-A6FE-494A1EFB515E, name="Basic data partition"
/dev/vda4 : start=    16582912, size=      193792, type=DE94BBA4-06D1-4D40-A16A-BFD50179D6AC, uuid=7BB90563-4BB7-4281-98EA-3FF4BCF1FCA5, attrs="RequiredPartition GUID:63"

[Fri Jan 26 root@fogclient ~]# cat /images/4KDisk-Base-Dev/d1.minimum.partitions
label: gpt
label-id: 9865AAFC-B984-4860-ACF5-4D6F2513747D
device: /dev/vda
unit: sectors
first-lba: 6
last-lba: 16777210
sector-size: 4096

/dev/vda1 : start=         256, size=       76800, type=C12A7328-F81F-11D2-BA4B-00A0C93EC93B, uuid=7C4743B5-7150-4672-B521-7B537528D7E7, name="EFI system partition", attrs="GUID:63"
/dev/vda2 : start=       77056, size=        4096, type=E3C9E316-0B5C-4DB8-817D-F92DF00215AE, uuid=F6216A84-0172-4445-B616-E36DFA20C731, name="Microsoft reserved partition", attrs="GUID:63"
/dev/vda3 : start=       81152, size=    16501760, type=EBD0A0A2-B9E5-4433-87C0-68B6B72699C7, uuid=B6DA06DC-A5D0-434C-A6FE-494A1EFB515E, name="Basic data partition"
/dev/vda4 : start=    16582912, size=      193792, type=DE94BBA4-06D1-4D40-A16A-BFD50179D6AC, uuid=7BB90563-4BB7-4281-98EA-3FF4BCF1FCA5, attrs="RequiredPartition GUID:63"

[Fri Jan 26 root@fogclient ~]# cat /images/4KDisk-Base-Dev/d1.fixed_size_partitions
1:2:4

[Fri Jan 26 root@fogclient ~]# cat /images/4KDisk-Base-Dev/d1.
d1.fixed_size_partitions  d1.minimum.partitions     d1.original.swapuuids     d1.shrunken.partitions
d1.mbr                    d1.original.fstypes       d1.partitions

[Fri Jan 26 root@fogclient ~]# cat /images/4KDisk-Base-Dev/d1.shrunken.partitions
label: gpt
label-id: 9865AAFC-B984-4860-ACF5-4D6F2513747D
device: /dev/vda
unit: sectors
first-lba: 6
last-lba: 16777210
sector-size: 4096

/dev/vda1 : start=         256, size=       76800, type=C12A7328-F81F-11D2-BA4B-00A0C93EC93B, uuid=7C4743B5-7150-4672-B521-7B537528D7E7, name="EFI system partition", attrs="GUID:63"
/dev/vda2 : start=       77056, size=        4096, type=E3C9E316-0B5C-4DB8-817D-F92DF00215AE, uuid=F6216A84-0172-4445-B616-E36DFA20C731, name="Microsoft reserved partition", attrs="GUID:63"
/dev/vda3 : start=       81152, size=    16501760, type=EBD0A0A2-B9E5-4433-87C0-68B6B72699C7, uuid=B6DA06DC-A5D0-434C-A6FE-494A1EFB515E, name="Basic data partition"
/dev/vda4 : start=    16582912, size=      193792, type=DE94BBA4-06D1-4D40-A16A-BFD50179D6AC, uuid=7BB90563-4BB7-4281-98EA-3FF4BCF1FCA5, attrs="RequiredPartition GUID:63"

gdisk -l /dev/sda
GPT fdisk (gdisk) version 1.0.8

Partition table scan:
  MBR: protective
  BSD: not present
  APM: not present
  GPT: present

Found valid GPT with protective MBR; using GPT.
Disk /dev/sda: 31246336 sectors, 119.2 GiB
Model: KLUDG4UHGC-B0E1
Sector size (logical/physical): 4096/4096 bytes
Disk identifier (GUID): 9865AAFC-B984-4860-ACF5-4D6F2513747D
Partition table holds up to 128 entries
Main partition table begins at sector 2 and ends at sector 5
First usable sector is 6, last usable sector is 31246330
Partitions will be aligned on 256-sector boundaries
Total free space is 14469877 sectors (55.2 GiB)

Number  Start (sector)    End (sector)  Size       Code  Name
   1             256           77055   300.0 MiB   EF00  EFI system partition
   2           77056           81151   16.0 MiB    0C01  Microsoft reserved ...
   3           81152        16582911   62.9 GiB    0700  Basic data partition
   4        16582912        16776703   757.0 MiB   2700

It’s a 128 GB drive, the image was a 64 GB drive, I expected it to expand to 128 GB

ntfsresize info on parts 4 and 3

 ntfsresize --info /dev/sda4
ntfsresize v2022.10.3 (libntfs-3g)
Device name        : /dev/sda4
NTFS volume version: 3.1
Cluster size       : 4096 bytes
Current volume size: 793772032 bytes (794 MB)
Current device size: 793772032 bytes (794 MB)
Checking filesystem consistency ...
100.00 percent completed
Accounting clusters ...
Space in use       : 14 MB (1.7%)
Collecting resizing constraints ...
You might resize at 13193216 bytes or 14 MB (freeing 780 MB).
Please make a test run using both the -n and -s options before real resizing!

[Fri Jan 26 root@fogclient ~]# ntfsresize --info /dev/sda3
ntfsresize v2022.10.3 (libntfs-3g)
Device name        : /dev/sda3
NTFS volume version: 3.1
Cluster size       : 4096 bytes
Current volume size: 67591208960 bytes (67592 MB)
Current device size: 67591208960 bytes (67592 MB)
Checking filesystem consistency ...
100.00 percent completed
Accounting clusters ...
Space in use       : 24382 MB (36.1%)
Collecting resizing constraints ...
You might resize at 24381546496 bytes or 24382 MB (freeing 43210 MB).
Please make a test run using both the -n and -s options before real resizing!

@Ceregon I’ve never messed with cloning a raid array. Anything can be done, but whether or not it’s going to work with built-in stuff is a different question.
I imagine you have vroc/vmd enabled in the bios on the machine where you’re deploying already. I’ve never got to play with Vroc but I’m familiar with it, just wasn’t able to convince management to buy me the stuff to try it a few years back.
My first guess is that /dev/md124 doesn’t exist because the raid volume doesn’t exist yet, but it sounds like you found that in a debug session on a host you’re trying to deploy too. So that’s probably out. But I just wonder if the VROC volume needs to be created beforehand to be deployed to, but I don’t have a full understanding of when that volume is made.

My next guess would be that a RAID array is a multiple disk system, so the image needs to be captured in multiple disk mode
Are you having different disk sizes for these RAID volumes? would capturing with multiple disk or dd be an option?
In theory a RAID is a single volume, and you may be able to capture it correctly and it sounds like you’ve found others in the forum that have done that?

Other possibility is the need for different VROC drivers in the bzImage kernel, but I feel like if that was the case, then you wouldn’t be able to see the disk at all when capturing.

You could also capture in debug mode and mount the windows drive before starting the capture to see if you can read stuff?
This is from part of a postdownload script that will mount the windows disk to the path /ntfs

. /usr/share/fog/lib/funcs.sh
mkdir -p /ntfs
getHardDisk
getPartitions $hd
for part in $parts; do
    umount /ntfs >/dev/null 2>&1
    fsTypeSetting "$part"
    case $fstype in
        ntfs)
            dots "Testing partition $part"
            ntfs-3g -o force,rw $part /ntfs
            ntfsstatus="$?"
            if [[ ! $ntfsstatus -eq 0 ]]; then
                echo "Skipped"
                continue
            fi
            if [[ ! -d /ntfs/windows && ! -d /ntfs/Windows && ! -d /ntfs/WINDOWS ]]; then
                echo "Not found"
                umount /ntfs >/dev/null 2>&1
                continue
            fi
            echo "Success"
            break
            ;;
        *)
            echo " * Partition $part not NTFS filesystem"
            ;;
    esac
done
if [[ ! $ntfsstatus -eq 0 ]]; then
    echo "Failed"
    debugPause
    handleError "Failed to mount $part ($0)\n    Args: $*"
fi
echo "Done"

Also, hot tip, once you’re in debug mode, you can run passwd and set a root password for that debug session. Then run ifconfig to get the ip. Then you can ssh into your debug session with ssh root@ip then put in the password you set when prompted. Then you can copy and paste this stuff and it’s a lot easier to copy the output or take screenshots.

Another possibilty could be using pre and post download scripts to fix the raid volume in the linux side, I found this information https://www.intel.com/content/dam/support/us/en/documents/memory-and-storage/linux-intel-vroc-userguide-333915.pdf but I didn’t dig into to that too much.

@limbooface Is this still an issue? If not, I’ll close this issue.

Did a debug session and ntfsresize with -c to check and --info shows partition 3 as resizable but it’s not being resized after imaging.

Running the image in deploy and seeing what it says after imaging and will see if there are any errors.

I fear this is going to be a 4kn drive alignment/resize issue.

@JJ-Fullmer I just took another look as I’ve been just maintaining a separate 4kn image and realized that the disk isn’t expanding after imaging.
I’ll do a debug session tomorrow and report back.

@sami-blackkite Looks like you’re on the dev branch, which is good. Updating is always a good first step in troubleshooting stuff like this in case it’s already fixed.
I’m not seeing anything in the commits (https://github.com/FOGProject/fogproject/commits/dev-branch) that actually looks related since the version you’re on, but sometimes a refresh still helps.

I took a look at the default settings of a new image and it’s close to what I use, only difference is I set compression to 11.

Are you familiar with a debug capture/deploy task? If not, simply check the ‘debug task’ checkbox when queuing up the capture task. It allows you to step through the capture process and you can catch any messages that might be telling us why it’s behaving odd. Just use the command fog once it’s booted up and ready to start the image and you’ll hit enter for each step. You can also set it up to watch it and step through over ssh if you want by, before running the fog command, getting the ip address with ifconfig and then setting a password with passwd then you can ssh into the debug session from your workstation with ssh root@ip.add.re.ss then the password you set. The password will only exist for that session on that machine. SSH just makes screenshotting and or copying any error messages to share here a bit easier.

@Lukaz You could try doing a post download script that takes the hostname you set for that fog host and put it in the /etc/hostname file.
I think there may be other locations that hostnamectl sets on newer linux builds, but this could still work.

I haven’t done a linux post download script but you can see some other examples in these posts, mostly windows based, but I’m sure we can apply the idea to linux without too much trouble. Just gotta mount the client disk in FOS (Fog operating system that you boot into over the network for the imaging process) after imaging is done and inject the fog hots hostname variable (not sure what the var name is off the top of my head) into /etc/hostname on the client. It’s a bit of work initially, but once setup should just work from then on.

https://forums.fogproject.org/topic/8889/fog-post-install-script-for-win-driver-injection?_=1682187993801

https://forums.fogproject.org/topic/7740/the-magical-mystical-fog-post-download-script

@sami-blackkite That’s the right option.
What version of FOG are you on?

@TanguyPSV I would suggest giving the dev-branch a try. See the “choosing a fog version” of this page.
https://docs.fogproject.org/en/latest/install-fog-server

Running the installer on the current server will upgrade your instance without losing the db. This will get you to the latest and greatest version.