FOG 1.5.9-RC2 incompatible with Windows 10 v2004 Partition Structure

[sudburr]

Uh oh.

The partition structure for a bare metal UEFI install of Windows 10 has changed dramatically with v2004.

FOG 1.5.9-RC2 (installed clean today at 3pm ET) no likey.

Microsoft Windows 10_v1903 64bit (10.0.18362.30)
Partition ### Type Size Offset
Partition 1 Recovery 529 MB 1024 KB
Partition 2 System 99 MB 530 MB
Partition 3 Reserved 16 MB 629 MB
Partition 4 Primary 1023 GB 645 MB

Microsoft Windows 10_v2004 64bit (10.0.19041.264) aka 20h1
Partition ### Type Size Offset
Partition 1 System 100 MB 1024 KB
Partition 2 Reserved 16 MB 101 MB
Partition 3 Primary 1023 GB 117 MB
Partition 4 Recovery 505 MB 1023 GB

It captures, something, but deployment well, that fails because v2004’s partition structure is not sized properly.

[george1421]

@sudburr said in FOG 1.5.9-RC2 incompatible with Windows 10 v2004 Partition Structure:

I would say this is the incorrect format with the recovery partition first

Microsoft Windows 10_v1903 64bit (10.0.18362.30)
Partition ### Type Size Offset
Partition 1 Recovery 529 MB 1024 KB
Partition 2 System 99 MB 530 MB
Partition 3 Reserved 16 MB 629 MB
Partition 4 Primary 1023 GB 645 MB

I would say this is a better disk structure in that the uefi boot partition is in the first position. d1p1. But I would not have the recovery partition as the last partition on the disk, if you have a choice on how is created move it ahead of the drive. I would say when we create the master image with MDT with the exception of the recovery partition this is how the disk structure looks.

Microsoft Windows 10_v2004 64bit (10.0.19041.264) aka 20h1
Partition ### Type Size Offset
Partition 1 System 100 MB 1024 KB
Partition 2 Reserved 16 MB 101 MB
Partition 3 Primary 1023 GB 117 MB
Partition 4 Recovery 505 MB 1023 GB

I would almost bet it has nothing to do with the partition location but instead some kind of hokus-pokus MS has done within the partitions themselves (like creating their own filesystem or something). FOG doesn’t really care about the content or location since it only copies blocks of data.

I guess its time that I load 2004 into MDT and spin up a new test golden image to see what is broken.

Either way thank you for letting the developers know that their summer break has now been canceled.

[Sebastian Roth]

@sudburr said in FOG 1.5.9-RC2 incompatible with Windows 10 v2004 Partition Structure:

It captures, something, but deployment well, that fails because v2004’s partition structure is not sized properly.

Thanks for reporting! Need more details. Contents of d1.partitions, d1.minimum.partitions and d1.fixed_size_partitions. Then as well a picture of the actual error - if there is any. From the sound of what you say I can imagine you mean it’s “just not” resizing this layout to the full disk size of a destination disk, right?

[sudburr]

[sudburr]

Original disk is 1 TB. Destination disk is 0.5 TB .

d1.partitions

label: gpt
label-id: FB16BC47-E9A4-4A26-9E00-495F842C2401
device: /dev/sda
unit: sectors
first-lba: 34
last-lba: 2147483614
sector-size: 512

/dev/sda1 : start=        2048, size=      204800, type=C12A7328-F81F-11D2-BA4B-00A0C93EC93B, uuid=DB56A1C9-1019-46FE-9C8A-8D561A3E1DE4, name="EFI system partition", attrs="GUID:63"
/dev/sda2 : start=      206848, size=       32768, type=E3C9E316-0B5C-4DB8-817D-F92DF00215AE, uuid=B1AC3238-84AA-410A-97AC-7C2683A132C5, name="Microsoft reserved partition", attrs="GUID:63"
/dev/sda3 : start=      239616, size=  2146203765, type=EBD0A0A2-B9E5-4433-87C0-68B6B72699C7, uuid=D6FFACFD-59C5-4AAA-995C-539E27EAD00A, name="Basic data partition"
/dev/sda4 : start=  2146445312, size=     1034240, type=DE94BBA4-06D1-4D40-A16A-BFD50179D6AC, uuid=A7345D7C-3D3F-44E9-B556-6F89394B0C76, attrs="RequiredPartition GUID:63"

=-=-=-=-=-

d1.minimum.partitions

label: gpt
label-id: FB16BC47-E9A4-4A26-9E00-495F842C2401
device: /dev/sda
unit: sectors
first-lba: 34
last-lba: 2147483614
sector-size: 512

/dev/sda1 : start=        2048, size=      204800, type=C12A7328-F81F-11D2-BA4B-00A0C93EC93B, uuid=DB56A1C9-1019-46FE-9C8A-8D561A3E1DE4, name="EFI system partition", attrs="GUID:63"
/dev/sda2 : start=      206848, size=       32768, type=E3C9E316-0B5C-4DB8-817D-F92DF00215AE, uuid=B1AC3238-84AA-410A-97AC-7C2683A132C5, name="Microsoft reserved partition", attrs="GUID:63"
/dev/sda3 : start=      239616, size=    20614004, type=EBD0A0A2-B9E5-4433-87C0-68B6B72699C7, uuid=D6FFACFD-59C5-4AAA-995C-539E27EAD00A, name="Basic data partition"
/dev/sda4 : start=  2146445312, size=     1034240, type=DE94BBA4-06D1-4D40-A16A-BFD50179D6AC, uuid=A7345D7C-3D3F-44E9-B556-6F89394B0C76, name="attrs=\x22RequiredPartition GUID:63"

=-=-=-=-=-

d1.fixed_size_partitions

1:2:4

[Sebastian Roth]

@sudburr Thanks! I am wondering if you really need that recovery partition (sda4)? This is kind of blocking for this partition layout to be shrinkable down to the size of your 500 GB disk. The start of sda4, sector 2146445312 is is not moved forward by FOS as we might cause an issue doing so.

A way to quickly fix this is using a partitioning too (probably even Windows disk management of your running system), shrink sda3 (C: in Windows) down to e.g. 400 GB, move the recovery partition forward as well (don’t think Windows disk management can do this but I am not sure) and then recapture the image.

[sudburr]

Shrinking partition 3, then moving partition 4, I can capture and deploy.

Going from a 110GB drive, to a 1024GB drive, partition 4 grows by roughly a factor of 10. The size of partition 4 is obviously based on % of disk space used on original instead of maintaining the fixed size that is intended.

Yes, a recovery partition is desired.

This is a workaround I can live with for now, until a fix is developed, but it puts a kink in the deployment process as we must now ensure a system has a drive that is no smaller than the original system the image was built on.

Makes me think that FOG is resizing the wrong partitions when capturing, though it says it has detected the correct fixed size partitions.

[Sebastian Roth]

@sudburr said in FOG 1.5.9-RC2 incompatible with Windows 10 v2004 Partition Structure:

The size of partition 4 is obviously based on % of disk space used on original instead of maintaining the fixed size that is intended.

Really?? Can you please take a picture of Windows disk management after the deploy and first reboot?

[sudburr]

@Sebastian-Roth The math bears it out. will check again

[sudburr]

Okay, that is peculiar. My test yesterday to a physical device resulted in the expanded partition 4. My test today to a VM did not. Partition 4 remained right-sized.

I don’t know when I’ll be able to do another physical test, but I will check again, somewhen.

[sudburr]

@Sebastian-Roth Okay, I wasn’t dreaming it.

Here’s the original on a 1024 GB drive after I’ve shrunk the partitions with GPARTED

And here’s the result immediately after going onto a 110 GB drive.

My original percentage math just happened to be a coincidence. But partition 4 is definitely not reproducing as intended.

[Sebastian Roth]

@sudburr Do I get this right? It does expand sda3 and sda4.

Is d1.fixed_size_partitions still set to 1:2:4 for this image?

[sudburr]

No.

cat d1.fixed_size_partitions
1:2

[Sebastian Roth]

@sudburr Did you manually edit the file or recapture the image or why is this changed? Just trying to make sense of this.

[sudburr]

It’s consistently inconsistent.

Today I created some more VMs. All with 64 GB drives. Essentially …

Capture VM1

cat d1.fixed_size_partitions
1:2

Capture VM1 a second time

cat d1.fixed_size_partitions
1:2:4

Capture VM2

cat d1.fixed_size_partitions
1:2:4

None will deploy to a drive smaller than the original 64 GB, though the data is only 12 GB uncompressed. It’s just a Windows install.

Looking at the 50 GB drive after a failed attempt to push the 64 GB image onto it. Diskpart reports a single 63 GB partition. wha?
Gnome Partition Editor shows the drive as 50 GB unallocated.

Dumping one of the 1:2:4 images onto a 2TB drive now; and it’s good.

So Partition 3, isn’t really resizing smaller when captured, and Partition 4 is sometimes not identified as fixed size.

[sudburr]

So far today, working with all new VMs again, things are looking good.

I made one change to the mastering process, to use Disk Management (diskmgmt.msc) within Windows to shrink the OS partition (to just 5GB free space) before sysprep and shutdown. It’s capturing partitions properly with fixed 1:2:4 so far.

Captured images are deploying properly to HDDs both larger and smaller than the original 64GB.

[Sebastian Roth]

@sudburr said:

Partition 4 is sometimes not identified as fixed size.

Is this something you can reproduce?

[sudburr]

Not reliably.

VM I’m working on now refuses to capture with partition 4 as fixed. It’s throwing the error during capture:

blkid: error: /dev/sda4: No such file or directory

[sudburr]

Okay, you’re going to love this. I only have one success to go on so far, but here’s my current working theory.

Failure scenario

1. Cold boot VM to Quick Inventory
2. Shutdown VM after the natural reboot after QI
3. Create Task
4. Cold boot VM into Task
5. Partition 4 is not recognized as fixed.

Success scenario

1. Cold boot VM to Quick Inventory
2. Pause VM after the natural reboot after QI
3. Create Task
4. Resume VM into Task
5. Partition 4 is recognized properly.

Testing further, but right now, Cold booting into the task appears to be the guilty party.

[Sebastian Roth]

@sudburr said in FOG 1.5.9-RC2 incompatible with Windows 10 v2004 Partition Structure:

Testing further, but right now, Cold booting into the task appears to be the guilty party.

Wow that would be a really nasty one. Please keep us posted.