Cannot deploy image after successful build. Is this an issue in windows packer template file. #278

iamtinu · 2024-11-25T11:12:00Z

The default template file for windows creates a successful build on an ubuntu 24 VM but the build image gives error that:

tar: This does not look like a tar archive

So I figured the error tar This does not look like a tar archive indicates that the downloaded image is not in the expected format (likely not a valid .tar.gz or .tgz archive).

Broken Pipe: The Cannot write to ‘-’ (Broken pipe) error occurs because tar exited prematurely due to the invalid archive format, breaking the data stream to dd.

Filesystem Not Found:
After the failure, no partitions or filesystems are found on /dev/sda, causing the deployment to fail entirely.

I modified default packer file to make a new image file and successfully deployed on a raid filesystem. Can be deployed on other disk layouts too.

Changes :
Filename Updated:
Changed the output filename to .tar.gz for better compatibility with MAAS.

tar.gz Compression:
Updated the post-processor section to compress the output using tar -czf, which produces a .tar.gz file MAAS can decompress

Dynamic Filesystem Mounting:
The mount command now dynamically attempts to mount common filesystem types (ntfs, ext4, vfat) for adding curtin hooks.

Added Output Sync:
Added explicit sync commands to ensure file operations are flushed before moving to the next step.

below is the changed file for successful image building and successful deployment.

windows.pkr.hcl.txt

uploaded instead of pasting code : rename .txt to windows.pkr.hcl. Recreate the image and reupload using

maas yourprofile boot-resources create
name='windows/windows-server'
title='Windows Server 2016'
architecture='amd64/generic'
filetype='ddtgz'
content@=2016.dd.tar.gz

update : dont change filetype while uploading image from ddtgz to tgz

Thank You

MggMuggins · 2024-11-25T15:09:54Z

Hi! Can you open a pull request so that we can more easily see the changes you made here? Thanks!

nics90 · 2024-11-25T15:42:30Z

Fixed some typos post which able to build the image:

packer {
  required_version = ">= 1.7.0"
  required_plugins {
    qemu = {
      version = "~> 1.0"
      source  = "github.com/hashicorp/qemu"
    }
  }
}


variable "headless" {
  type        = bool
  default     = false
  description = "Whether VNC viewer should not be launched."
}

variable "disk_size" {
  type    = string
  default = "32G"
}

variable "iso_path" {
  type    = string
  default = ""
}

variable "ovmf_suffix" {
  type    = string
  default = "_4M"
}

variable "filename" {
  type        = string
  default     = "windows.dd.tar.gz"
  description = "The filename of the tarball to produce"
}

variable "is_vhdx" {
  type        = bool
  default     = false
  description = "Whether we are building using a VHDX disk."
}

variable "use_tpm" {
  type        = string
  default     = "false"
  description = "Whether we are building using a virtual TPM device."
}

variable "timeout" {
  type    = string
  default = "1h"
}

locals {
  baseargs = [
    ["-cpu", "host"],
    ["-serial", "stdio"],
    ["-drive", "if=pflash,format=raw,id=ovmf_code,readonly=on,file=/usr/share/OVMF/OVMF_CODE${var.ovmf_suffix}.ms.fd"],
    ["-drive", "if=pflash,format=raw,id=ovmf_vars,file=OVMF_VARS.fd"],
    ["-drive", "file=output-windows_builder/packer-windows_builder,format=raw"],
    ["-cdrom", "${var.iso_path}"],
    ["-drive", "file=drivers.iso,media=cdrom,index=3"],
    ["-boot", "d"]
  ]
  tpmargs = [
    ["-chardev", "socket,id=chrtpm,path=/tmp/swtpm/swtpm-sock"],
    ["-tpmdev", "emulator,id=tpm0,chardev=chrtpm"],
    ["-device", "tpm-tis,tpmdev=tpm0"]
  ]
}

source "qemu" "windows_builder" {
  accelerator       = "kvm"
  boot_command      = [""]
  boot_wait         = "2s"
  communicator      = "none"
  disk_interface    = "ide"
  disk_image        = "${var.is_vhdx}"
  disk_size         = "${var.disk_size}"
  floppy_files      = ["./http/Autounattend.xml", "./http/logon.ps1", "./http/rh.cer"]
  floppy_label      = "flop"
  format            = "raw"
  headless          = "${var.headless}"
  http_directory    = "http"
  iso_checksum      = "none"
  iso_url           = "${var.iso_path}"
  machine_type      = "q35"
  memory            = "4096"
  cpus              = "2"
  net_device        = "e1000"
  qemuargs          = concat(local.baseargs, (var.use_tpm == "yes" ? local.tpmargs : []))
  shutdown_timeout  = "${var.timeout}"
  vnc_bind_address  = "0.0.0.0"
}

build {
  sources = ["source.qemu.windows_builder"]

  post-processor "shell-local" {
    inline = [
      "echo 'Syncing output-windows_builder/packer-windows_builder...'",
      "sync -f output-windows_builder/packer-windows_builder",
      "IMG_FMT=raw",
      "source scripts/setup-nbd",
      "TMP_DIR=$(mktemp -d /tmp/packer-maas-XXXX)",
      "echo 'Adding curtin-hooks to image...'",
      "mount -t ntfs $${nbd}p3 $TMP_DIR || mount -t ext4 $${nbd}p3 $TMP_DIR || mount -t vfat $${nbd}p3 $TMP_DIR",
      "mkdir -p $TMP_DIR/curtin",
      "cp ./curtin/* $TMP_DIR/curtin/",
      "sync -f $TMP_DIR/curtin",
      "umount $TMP_DIR",
      "qemu-nbd -d $nbd",
      "rmdir $TMP_DIR",
      "tar -czf ${var.filename} -C output-windows_builder packer-windows_builder",
      "echo 'Image created: ${var.filename}'"
    ]
    inline_shebang = "/bin/bash -e"
  }

  post-processor "compress" {
    output = "${var.filename}"
  }
}

The installation finished but OS failed to come up with below errors:

alanbach · 2024-11-25T18:25:00Z

Hello Everyone! Please note that Windows images are not tar.gz. They are DD images and need to be imported accordingly in MAAS. Please see the README file for an example also this:

$ maas ${PROFILE} boot-resources create name='windows/windows-11' title='Windows 11' architecture='amd64/generic' filetype=ddgz content@=windows11.dd.gz

nics90 · 2024-11-26T06:08:12Z

@alanbach : I created dd image from the template in the repo, still after deploying getting the same error:

'Unable to mount root fs on unknown block'

iamtinu · 2024-11-26T07:54:33Z

windows.pkr.hcl.txt

I was able to successfully build image and tested successful deployment on a proxmox VM.

Attaching the modified packer file in txt format as github allow limited file extensions for upload.

But I think there is some issue on baremetal servers. My two dedicated servers wont go past windows logo. probably bios setting but I doubt it. I have uefi boot enabled, secure boot disabled, and disabled TPM on server and everything works fine in VM. but different story on a baremetal.

iamtinu · 2024-11-26T07:58:49Z

Hello Everyone! Please note that Windows images are not tar.gz. They are DD images and need to be imported accordingly in MAAS. Please see the README file for an example also this:
$ maas ${PROFILE} boot-resources create name='windows/windows-11' title='Windows 11' architecture='amd64/generic' filetype=ddgz content@=windows11.dd.gz

@alanbach You can upload successfully but you will run into error "This does not look like a tar archive" while deployment.

Hence the whole thread.

iamtinu · 2024-11-26T08:00:09Z

@alanbach : I created dd image from the template in the repo, still after deploying getting the same error:

'Unable to mount root fs on unknown block'

@nics90 try the changes I mentioned in the notepad file from my comments. and use that packer file to generate image for deployment

iamtinu · 2024-11-26T08:12:54Z

Hi! Can you open a pull request so that we can more easily see the changes you made here? Thanks!

@MggMuggins Updated the main thread with a notepad file. rename it and rebuild the image and try deployment then.

nics90 · 2024-11-26T10:59:54Z

@iamtinu : I uploaded the image using below command:

maas admin boot-resources create name='windows/win2k19' title='Windows Server 2019' architecture='amd64/generic' filetype='ddtgz' content@=/home/devops/windows2019.dd.tar.gz

and I'm using KVM with maas latest version to commission the machine, but got same error.

iamtinu · 2024-11-27T10:12:08Z

I cant test on KVM I am using another proxmox server and two baremetal servers. worked for me.

nics90 · 2024-11-27T10:31:17Z

Did it work for you on baremetal ? Since I also tried on baremetal, operating system got installed but failed to come up :

dellsrv3.maas-installation-output-2024-11-27.log

iamtinu · 2024-11-27T13:20:20Z

Did it work for you on baremetal ? Since I also tried on baremetal, operating system got installed but failed to come up :

dellsrv3.maas-installation-output-2024-11-27.log

on baremetal my OS installs. that specific error is related to uefi I think. check TPM disabled, CSM disabled, uefi on and correct boot partition is marked as second boot option after setting network boot as first. I think you already installed OS just boot problem.
If that not works. release the server and quick erase disk . commission the machine again, then deploy.

in your previous logs I can also see PERC H755 Front so ensure drivers for the disk controller (PERC H755 Front) are loaded. Check with:
lspci -k | grep -A 2 PERC

If the drivers for PERC H755 are missing, rebuild the initramfs:
chroot /mnt
update-initramfs -u

Missing Bootloader or EFI Files. The EFI partition (/dev/sda1) was wiped and likely not re-populated with GRUB/bootloader files.
Without these files, the system cannot boot.

github-actions · 2024-12-28T01:56:50Z

This issue is stale because it has been open for 30 days with no activity.

tdtmusic2 · 2025-01-09T15:08:56Z

Hi everyone. In my case (baremetal), using the built-in packer windows template script resulted in a failed deployment in maas, it didn't even start to download the image and just gave the error This does not look like a tar archive. After replacing the template script with the one in the .txt file above, recreating the image, uploading it to maas and retrying the deployment it actually downloaded the image this time, but then I got these and a failed deployment again:

writing image to disk {'type': 'dd-tgz', 'uri': 'http://172.17.17.105:5248/images/87b8187/windows/amd64/generic/windowsserver2022/uploaded/root-dd'}, sda
        Running command ['sh', '-c', 'wget "$1" --progress=dot:mega -O - |tar -xOzf -| dd bs=4M of="$2"', '--', 'http://172.17.17.105:5248/images/87b8187/windows/amd64/generic/windowsserver2022/uploaded/root-dd', '/dev/sda'] with allowed return codes [0] (capture=False)
        --2025-01-09 13:10:27--  http://172.17.17.105:5248/images/87b8187/windows/amd64/generic/windowsserver2022/uploaded/root-dd
        Connecting to 172.17.17.105:5248... connected.
        HTTP request sent, awaiting response... 200 OK
        Length: 6645865672 (6.2G) [application/octet-stream]
        Saving to: ‘STDOUT’
        
             0K ....tar: This does not look like a tar archive
        tar: Skipping to next header
        .tar: Exiting with failure status due to previous errors
                                                          0% 99.1M=0.004s
        
        
        Cannot write to ‘-’ (Broken pipe).
        0+0 records in
        0+0 records out
        0 bytes copied, 0.0227721 s, 0.0 kB/s
        Running command ['partprobe', '/dev/sda'] with allowed return codes [0] (capture=False)
        Running command ['udevadm', 'trigger', '/dev/sda'] with allowed return codes [0] (capture=False)
        Running command ['udevadm', 'settle'] with allowed return codes [0] (capture=False)
        TIMED udevadm_settle(): 0.091
        Running command ['vgchange', '--activate=y'] with allowed return codes [0] (capture=True)
        Running command ['udevadm', 'settle'] with allowed return codes [0] (capture=False)
        TIMED udevadm_settle(): 0.019
        Running command ['udevadm', 'trigger', '/dev/sda'] with allowed return codes [0] (capture=False)
        Running command ['udevadm', 'settle'] with allowed return codes [0] (capture=False)
        TIMED udevadm_settle(): 0.042
        Running command ['vgchange', '--activate=y'] with allowed return codes [0] (capture=True)
        Running command ['udevadm', 'settle'] with allowed return codes [0] (capture=False)
        TIMED udevadm_settle(): 0.011
        Running command ['udevadm', 'trigger', '/dev/sda'] with allowed return codes [0] (capture=False)
        Running command ['udevadm', 'settle'] with allowed return codes [0] (capture=False)
        TIMED udevadm_settle(): 0.043
        Running command ['vgchange', '--activate=y'] with allowed return codes [0] (capture=True)
        Running command ['udevadm', 'settle'] with allowed return codes [0] (capture=False)
        TIMED udevadm_settle(): 0.012
        Searching for filesystem on ['sda'] containing one of: ['curtin', 'system-data/var/lib/snapd', 'snaps']
        Running command ['lsblk', '--noheadings', '--bytes', '--pairs', '--output=ALIGNMENT,DISC-ALN,DISC-GRAN,DISC-MAX,DISC-ZERO,FSTYPE,GROUP,KNAME,LABEL,LOG-SEC,MAJ:MIN,MIN-IO,MODE,MODEL,MOUNTPOINT,NAME,OPT-IO,OWNER,PHY-SEC,RM,RO,ROTA,RQ-SIZE,SIZE,STATE,TYPE,UUID', '/dev/sda'] with allowed return codes [0] (capture=True)
        Known partitions []
        TIMED BLOCK_META: 2.539
        finish: cmd-install/stage-partitioning/builtin/cmd-block-meta: FAIL: curtin command block-meta
        Traceback (most recent call last):
          File "/curtin/curtin/commands/main.py", line 202, in main
            ret = args.func(args)
          File "/curtin/curtin/log.py", line 97, in wrapper
            return log_time("TIMED %s: " % msg, func, *args, **kwargs)
          File "/curtin/curtin/log.py", line 79, in log_time
            return func(*args, **kwargs)
          File "/curtin/curtin/commands/block_meta.py", line 120, in block_meta
            return meta_simple(args)
          File "/curtin/curtin/commands/block_meta.py", line 2321, in meta_simple
            rootdev = write_image_to_disk(dd_images[0], devname)
          File "/curtin/curtin/commands/block_meta.py", line 182, in write_image_to_disk
            return block.get_root_device([devname], paths=paths)
          File "/curtin/curtin/block/__init__.py", line 761, in get_root_device
            raise ValueError(
        ValueError: Did not find any filesystem on ['sda'] that contained one of ['curtin', 'system-data/var/lib/snapd', 'snaps']
        Did not find any filesystem on ['sda'] that contained one of ['curtin', 'system-data/var/lib/snapd', 'snaps']
        
Stderr: ''

alanbach · 2025-01-09T15:49:36Z

Hello @tdtmusic2 Looking at the above logs, the deployment is failing due to curtin directory missing in the image. I do see your image was uploaded as a dd image which is correct, however the missing curtin directory usually means the build process was never fully completed.

Would you please re-run the build process, increase the timeout from the default 1 hour and watch the process closely to see if you get errors / interruptions at any stage?

tdtmusic2 · 2025-01-09T16:32:04Z

Hello @tdtmusic2 Looking at the above logs, the deployment is failing due to curtin directory missing in the image. I do see your image was uploaded as a dd image which is correct, however the missing curtin directory usually means the build process was never fully completed.

Would you please re-run the build process, increase the timeout from the default 1 hour and watch the process closely to see if you get errors / interruptions at any stage?

Hello. This is the output from the building process, it seems fine to me since it reports success but maybe I'm missing something:

rm -f -rf output-* windows.dd.gz http/Autounattend.xml \
drivers.iso windows_builder.img OVMF_VARS.fd
cp -v /usr/share/OVMF/OVMF_VARS_4M.ms.fd OVMF_VARS.fd
'/usr/share/OVMF/OVMF_VARS_4M.ms.fd' -> 'OVMF_VARS.fd'
mkisofs -o drivers.iso -V 'Extra Drivers' -input-charset utf-8 drivers
Total translation table size: 0
Total rockridge attributes bytes: 0
Total directory bytes: 2048
Path table size(bytes): 22
Max brk space used 0
176 extents written (0 MB)
sed s#@VERSION@#"Windows Server 2022 SERVERSTANDARD"#g http/Autounattend.xml.ISO.template > http/Autounattend.xml
packer init . && packer build -var iso_path=/mnt/iso/SERVER_EVAL_x64FRE_en-us.iso \
                -var headless=false \
                -var ovmf_suffix=_4M \
                -var is_vhdx=false \
                -var timeout=1h \
                -var use_tpm= \
                windows.pkr.hcl
Warning: A checksum of 'none' was specified. Since ISO files are so big,
a checksum is highly recommended.

  on windows.pkr.hcl line 74:
  (source code not available)


qemu.windows_builder: output will be in this color.

==> qemu.windows_builder: Retrieving ISO
==> qemu.windows_builder: Trying /mnt/iso/SERVER_EVAL_x64FRE_en-us.iso
==> qemu.windows_builder: Trying /mnt/iso/SERVER_EVAL_x64FRE_en-us.iso
==> qemu.windows_builder: /mnt/iso/SERVER_EVAL_x64FRE_en-us.iso => /mnt/iso/SERVER_EVAL_x64FRE_en-us.iso
==> qemu.windows_builder: Creating floppy disk...
    qemu.windows_builder: Copying files flatly from floppy_files
    qemu.windows_builder: Copying file: ./http/Autounattend.xml
    qemu.windows_builder: Copying file: ./http/logon.ps1
    qemu.windows_builder: Copying file: ./http/rh.cer
    qemu.windows_builder: Done copying files from floppy_files
    qemu.windows_builder: Collecting paths from floppy_dirs
    qemu.windows_builder: Resulting paths from floppy_dirs : []
    qemu.windows_builder: Done copying paths from floppy_dirs
    qemu.windows_builder: Copying files from floppy_content
    qemu.windows_builder: Done copying files from floppy_content
==> qemu.windows_builder: Starting HTTP server on port 8729
    qemu.windows_builder: No communicator is set; skipping port forwarding setup.
==> qemu.windows_builder: Looking for available port between 5900 and 6000 on 0.0.0.0
==> qemu.windows_builder: Starting VM, booting from CD-ROM
==> qemu.windows_builder: Overriding default Qemu arguments with qemuargs template option...
==> qemu.windows_builder: Waiting 2s for boot...
==> qemu.windows_builder: Connecting to VM via VNC (0.0.0.0:5908)
==> qemu.windows_builder: Typing the boot commands over VNC...
    qemu.windows_builder: No communicator is configured -- skipping StepWaitGuestAddress
==> qemu.windows_builder: Waiting for shutdown...
==> qemu.windows_builder: Running post-processor:  (type shell-local)
==> qemu.windows_builder (shell-local): Running local shell script: /tmp/packer-shell2174978315
    qemu.windows_builder (shell-local): Syncing output-windows_builder/packer-windows_builder...
    qemu.windows_builder (shell-local): output-windows_builder/packer-windows_builder
    qemu.windows_builder (shell-local): Loading nbd...
    qemu.windows_builder (shell-local): Using /dev/nbd0
    qemu.windows_builder (shell-local): Binding image to /dev/nbd0...
    qemu.windows_builder (shell-local): /dev/nbd0 disconnected
    qemu.windows_builder (shell-local): Waiting for partitions to be created...
    qemu.windows_builder (shell-local): Adding curtin-hooks to image...
    qemu.windows_builder (shell-local): /dev/nbd0 disconnected
    qemu.windows_builder (shell-local): Image created: windows.dd.tar.gz
==> qemu.windows_builder: Running post-processor:  (type compress)
==> qemu.windows_builder (compress): Using pgzip compression with 4 cores for windows.dd.tar.gz
==> qemu.windows_builder (compress): Tarring windows.dd.tar.gz with pgzip
==> qemu.windows_builder (compress): Archive windows.dd.tar.gz completed
Build 'qemu.windows_builder' finished after 27 minutes 33 seconds.

==> Wait completed after 27 minutes 33 seconds

==> Builds finished. The artifacts of successful builds are:
--> qemu.windows_builder: VM files in directory: output-windows_builder
--> qemu.windows_builder: VM files in directory: output-windows_builder
--> qemu.windows_builder: compressed artifacts in: windows.dd.tar.gz
rm http/Autounattend.xml

alanbach · 2025-01-12T20:44:58Z

Thank you @tdtmusic2! This build looks clean however there are circumstances where deleting and re-importing an image in MAAS does not actually refresh the images cached on Rack Controllers if the same name was used.

Do you mind deleting the image, then uploading it using a different name (name= parameter)? Add a suffix of some sort to make it unique then try the deployment again.

tdtmusic2 · 2025-01-19T20:15:41Z

Thank you @tdtmusic2! This build looks clean however there are circumstances where deleting and re-importing an image in MAAS does not actually refresh the images cached on Rack Controllers if the same name was used.

Do you mind deleting the image, then uploading it using a different name (name= parameter)? Add a suffix of some sort to make it unique then try the deployment again.

Hi Alan. Did that, but unfortunatelly it's the same result, not sure if the logs are identical but here's the output:

2025-01-13 10:15:19 (17.1 MB/s) - written to stdout [6509127766/6509127766]

0+1369175 records in
0+1369175 records out
34359738368 bytes (34 GB, 32 GiB) copied, 373.136 s, 92.1 MB/s
Running command ['partprobe', '/dev/sda'] with allowed return codes [0] (capture=False)
Warning: Not all of the space available to /dev/sda appears to be used, you can fix the GPT to use all of the space (an extra 182829696 blocks) or continue with the current setting? 
Running command ['udevadm', 'trigger', '/dev/sda'] with allowed return codes [0] (capture=False)
Running command ['udevadm', 'settle'] with allowed return codes [0] (capture=False)
TIMED udevadm_settle(): 0.260
Running command ['vgchange', '--activate=y'] with allowed return codes [0] (capture=True)
Running command ['udevadm', 'settle'] with allowed return codes [0] (capture=False)
TIMED udevadm_settle(): 0.022
Running command ['udevadm', 'trigger', '/dev/sda'] with allowed return codes [0] (capture=False)
Running command ['udevadm', 'settle'] with allowed return codes [0] (capture=False)
TIMED udevadm_settle(): 0.096
Running command ['vgchange', '--activate=y'] with allowed return codes [0] (capture=True)
Running command ['udevadm', 'settle'] with allowed return codes [0] (capture=False)
TIMED udevadm_settle(): 0.020
Running command ['udevadm', 'trigger', '/dev/sda'] with allowed return codes [0] (capture=False)
Running command ['udevadm', 'settle'] with allowed return codes [0] (capture=False)
TIMED udevadm_settle(): 0.091
Running command ['vgchange', '--activate=y'] with allowed return codes [0] (capture=True)
Running command ['udevadm', 'settle'] with allowed return codes [0] (capture=False)
TIMED udevadm_settle(): 0.014
Searching for filesystem on ['sda'] containing one of: ['curtin', 'system-data/var/lib/snapd', 'snaps']
Running command ['lsblk', '--noheadings', '--bytes', '--pairs', '--output=ALIGNMENT,DISC-ALN,DISC-GRAN,DISC-MAX,DISC-ZERO,FSTYPE,GROUP,KNAME,LABEL,LOG-SEC,MAJ:MIN,MIN-IO,MODE,MODEL,MOUNTPOINT,NAME,OPT-IO,OWNER,PHY-SEC,RM,RO,ROTA,RQ-SIZE,SIZE,STATE,TYPE,UUID', '/dev/sda'] with allowed return codes [0] (capture=True)
Known partitions ['sda1', 'sda2', 'sda3', 'sda4']
Running command ['mount', '/dev/sda1', '/tmp/tmpjveelamx'] with allowed return codes [0] (capture=False)
Running command ['umount', '/tmp/tmpjveelamx'] with allowed return codes [0] (capture=False)
Running command ['mount', '/dev/sda2', '/tmp/tmpjveelamx'] with allowed return codes [0] (capture=False)
mount: /tmp/tmpjveelamx: wrong fs type, bad option, bad superblock on /dev/sda2, missing codepage or helper program, or other error.
Running command ['mount', '/dev/sda3', '/tmp/tmpjveelamx'] with allowed return codes [0] (capture=False)
Found path 'curtin' on device '/dev/sda3'
Running command ['umount', '/tmp/tmpjveelamx'] with allowed return codes [0] (capture=False)
Running command ['mount', '/dev/sda4', '/tmp/tmpjveelamx'] with allowed return codes [0] (capture=False)
Running command ['umount', '/tmp/tmpjveelamx'] with allowed return codes [0] (capture=False)
Running command ['mount', '/dev/sda3', '/tmp/tmpeqn8vhm5/target'] with allowed return codes [0] (capture=False)
TIMED BLOCK_META: 376.005
finish: cmd-install/stage-partitioning/builtin/cmd-block-meta: SUCCESS: curtin command block-meta
start: cmd-install/stage-network/builtin/cmd-net-meta: curtin command net-meta
net-meta mode is 'custom'.  devices=[]
writing to file /tmp/tmpeqn8vhm5/state/network_config with network config: network:
  config:
  - id: eno1
    mac_address: 80:18:44:f0:79:ac
    mtu: 1500
    name: eno1
    subnets:
    - address: 172.17.17.223/24
      dns_nameservers:
      - 172.17.17.105
      dns_search:
      - maas
      gateway: 172.17.17.1
      type: static
    type: physical
  - id: eno2
    mac_address: 80:18:44:f0:79:ad
    mtu: 1500
    name: eno2
    subnets:
    - type: manual
    type: physical
  - id: eno3
    mac_address: 80:18:44:f0:79:ae
    mtu: 1500
    name: eno3
    subnets:
    - type: manual
    type: physical
  - id: eno4
    mac_address: 80:18:44:f0:79:af
    mtu: 1500
    name: eno4
    subnets:
    - type: manual
    type: physical
  - id: enp193s0f0
    mac_address: 34:80:0d:03:e7:10
    mtu: 1500
    name: enp193s0f0
    subnets:
    - type: manual
    type: physical
  - id: enp193s0f1
    mac_address: 34:80:0d:03:e7:11
    mtu: 1500
    name: enp193s0f1
    subnets:
    - type: manual
    type: physical
  - id: enp193s0f2
    mac_address: 34:80:0d:03:e7:12
    mtu: 1500
    name: enp193s0f2
    subnets:
    - type: manual
    type: physical
  - id: enp193s0f3
    mac_address: 34:80:0d:03:e7:13
    mtu: 1500
    name: enp193s0f3
    subnets:
    - type: manual
    type: physical
  - address:
    - 172.17.17.105
    search:
    - maas
    type: nameserver
  version: 1

finish: cmd-install/stage-network/builtin/cmd-net-meta: SUCCESS: curtin command net-meta
start: cmd-install/stage-extract/builtin/cmd-extract: curtin command extract
Installing sources: [{'type': 'dd-tgz', 'uri': 'http://172.17.17.105:5248/images/0469827/windows/amd64/generic/windowsserver2022standard/uploaded/root-dd'}] to target at /tmp/tmpeqn8vhm5/target
start: cmd-install/stage-extract/builtin/cmd-extract: acquiring and extracting image from http://172.17.17.105:5248/images/0469827/windows/amd64/generic/windowsserver2022standard/uploaded/root-dd
finish: cmd-install/stage-extract/builtin/cmd-extract: SUCCESS: acquiring and extracting image from http://172.17.17.105:5248/images/0469827/windows/amd64/generic/windowsserver2022standard/uploaded/root-dd
Applying write_files from config.
finish: cmd-install/stage-extract/builtin/cmd-extract: SUCCESS: curtin command extract
start: cmd-install/stage-curthooks/builtin/cmd-curthooks: curtin command curthooks
running /tmp/tmpeqn8vhm5/target/curtin/curtin-hooks
Running command ['/tmp/tmpeqn8vhm5/target/curtin/curtin-hooks'] with allowed return codes [0] (capture=False)
finish: cmd-install/stage-curthooks/builtin/cmd-curthooks: SUCCESS: curtin command curthooks
start: cmd-install/stage-hook/builtin/cmd-hook: curtin command hook
Finalizing /tmp/tmpeqn8vhm5/target
running /tmp/tmpeqn8vhm5/target/curtin/finalize
Running command ['/tmp/tmpeqn8vhm5/target/curtin/finalize'] with allowed return codes [0] (capture=False)
Traceback (most recent call last):
  File "/tmp/tmpeqn8vhm5/target/curtin/finalize.py", line 144, in <module>
    curthooks()
  File "/tmp/tmpeqn8vhm5/target/curtin/finalize.py", line 88, in curthooks
    cloudbaseinit = get_cloudbaseinit_dir(target)
  File "/tmp/tmpeqn8vhm5/target/curtin/finalize.py", line 82, in get_cloudbaseinit_dir
    raise ValueError("Failed to find cloudbase-init install destination")
ValueError: Failed to find cloudbase-init install destination
finish: cmd-install/stage-hook/builtin/cmd-hook: FAIL: curtin command hook
Traceback (most recent call last):
  File "/curtin/curtin/commands/main.py", line 202, in main
    ret = args.func(args)
  File "/curtin/curtin/commands/hook.py", line 26, in hook
    curtin.util.run_hook_if_exists(args.target, "finalize")
  File "/curtin/curtin/util.py", line 984, in run_hook_if_exists
    subp([target_hook])
  File "/curtin/curtin/util.py", line 280, in subp
    return _subp(*args, **kwargs)
  File "/curtin/curtin/util.py", line 144, in _subp
    raise ProcessExecutionError(stdout=out, stderr=err,
curtin.util.ProcessExecutionError: Unexpected error while running command.
Command: ['/tmp/tmpeqn8vhm5/target/curtin/finalize']
Exit code: 1
Reason: -
Stdout: ''
Stderr: ''
Unexpected error while running command.
Command: ['/tmp/tmpeqn8vhm5/target/curtin/finalize']
Exit code: 1
Reason: -
Stdout: ''
Stderr: ''
curtin: Installation failed with exception: Unexpected error while running command.
Command: ['curtin', 'hook']
Exit code: 3
Reason: -
Stdout: start: cmd-install/stage-hook/builtin/cmd-hook: curtin command hook
        Finalizing /tmp/tmpeqn8vhm5/target
        running /tmp/tmpeqn8vhm5/target/curtin/finalize
        Running command ['/tmp/tmpeqn8vhm5/target/curtin/finalize'] with allowed return codes [0] (capture=False)
        Traceback (most recent call last):
          File "/tmp/tmpeqn8vhm5/target/curtin/finalize.py", line 144, in <module>
            curthooks()
          File "/tmp/tmpeqn8vhm5/target/curtin/finalize.py", line 88, in curthooks
            cloudbaseinit = get_cloudbaseinit_dir(target)
          File "/tmp/tmpeqn8vhm5/target/curtin/finalize.py", line 82, in get_cloudbaseinit_dir
            raise ValueError("Failed to find cloudbase-init install destination")
        ValueError: Failed to find cloudbase-init install destination
        finish: cmd-install/stage-hook/builtin/cmd-hook: FAIL: curtin command hook
        Traceback (most recent call last):
          File "/curtin/curtin/commands/main.py", line 202, in main
            ret = args.func(args)
          File "/curtin/curtin/commands/hook.py", line 26, in hook
            curtin.util.run_hook_if_exists(args.target, "finalize")
          File "/curtin/curtin/util.py", line 984, in run_hook_if_exists
            subp([target_hook])
          File "/curtin/curtin/util.py", line 280, in subp
            return _subp(*args, **kwargs)
          File "/curtin/curtin/util.py", line 144, in _subp
            raise ProcessExecutionError(stdout=out, stderr=err,
        curtin.util.ProcessExecutionError: Unexpected error while running command.
        Command: ['/tmp/tmpeqn8vhm5/target/curtin/finalize']
        Exit code: 1
        Reason: -
        Stdout: ''
        Stderr: ''
        Unexpected error while running command.
        Command: ['/tmp/tmpeqn8vhm5/target/curtin/finalize']
        Exit code: 1
        Reason: -
        Stdout: ''
        Stderr: ''
        
Stderr: ''

alanbach · 2025-01-20T21:41:29Z

Hello @tdtmusic2 This error is a different issue! It looks like for some reason the image is missing cloud-base-init:

ValueError: Failed to find cloudbase-init install destination

Note that Internet connectivity is needed for the build process to be able to download all the needed artifacts. If the build VM does have internet, I recommend either watching the build process in headless = false mode or connecting to the VNC port to see what happens during the build. This might help catching the issue.

tdtmusic2 · 2025-01-23T10:38:52Z

Hello @tdtmusic2 This error is a different issue! It looks like for some reason the image is missing cloud-base-init:
ValueError: Failed to find cloudbase-init install destination
Note that Internet connectivity is needed for the build process to be able to download all the needed artifacts. If the build VM does have internet, I recommend either watching the build process in headless = false mode or connecting to the VNC port to see what happens during the build. This might help catching the issue.

Hello. I actually needed to set headless to true so I can get a vnc port to connect to, so I did just that. The Windows setup starts, completes the first phase, reboots, then does the preparing devices thing and reboots again. And then it boots all the way into Windows, Administratot login. A powershell instance starts and downloads some things, then a sysprep prompt shows up, with "Processing generalize phase Sysprep plugins" and after that the vm shuts down. Everything seems fine to me for this part. Also in the terminal all seems good. Here are some screenshots of the process.

tdtmusic2 · 2025-01-23T12:37:38Z

Also, no idea if related somehow, but I've noticed that every time I deploy the machine with Windows, when the image is downloaded it goes flawlessly with 100MB/s right up to 98%. At 98% it drops to around 800KB/s for some reason and it doesn't improve till the end. Weird.

alanbach · 2025-01-23T16:23:00Z

Hello @tdtmusic2 These actually look pretty clean.

Do you want to delete the old image from MAAS and upload this new one and re-test? Please do use a different name (name=windowssever2022v2 or something like that) to make sure the image is actually replaced on the Rack Controllers. Generally it is always a good practice to use unique names in the images.

Also the dd slowness at the end is a normal behavior.

Another question: From the logs, this seems to be Windows Server 2022 (please correct me if I am wrong). What ISO did you use for this? (Just curious)

tdtmusic2 · 2025-01-23T17:48:34Z

Hello @tdtmusic2 These actually look pretty clean.

Do you want to delete the old image from MAAS and upload this new one and re-test? Please do use a different name (name=windowssever2022v2 or something like that) to make sure the image is actually replaced on the Rack Controllers. Generally it is always a good practice to use unique names in the images.

Also the dd slowness at the end is a normal behavior.

Another question: From the logs, this seems to be Windows Server 2022 (please correct me if I am wrong). What ISO did you use for this? (Just curious)

I did that but unfortunatelly it ended prettty much the same way, Failed to find cloudbase-init install destination. You're right, it's a Windows Server 2022, the iso is the official one from Microsoft's evaluation site. I'll try a different version, like 2019, to see if it's the same result.

Later edit: Exactly the same error with a 2025 version.

tdtmusic2 · 2025-01-25T20:01:13Z

Hello @tdtmusic2 These actually look pretty clean.

Do you want to delete the old image from MAAS and upload this new one and re-test? Please do use a different name (name=windowssever2022v2 or something like that) to make sure the image is actually replaced on the Rack Controllers. Generally it is always a good practice to use unique names in the images.

Also the dd slowness at the end is a normal behavior.

Another question: From the logs, this seems to be Windows Server 2022 (please correct me if I am wrong). What ISO did you use for this? (Just curious)

No idea what could've caused this but trying on a different physical server worked, I already tested all versions of Windows Server since 2016. There is one small issue though, MaaS remains in Deploying (Rebooting) status sometimes and after a while it goes into Failed deployment, even if the OS is installed and running fine for a while. I don't think this breaks anything but I don't think it should happen.

Can you please give me some more details about "Also the dd slowness at the end is a normal behavior."?

alanbach · 2025-01-27T01:18:53Z

Hello @tdtmusic2 Glad to hear finally some progress! Regarding the machines getting into failed deployed even though the actual deployment was fine, I'd say check the cloud-init logs they should have hints as to why the post process failed.

Another Question: Is the machine hostname, IP address etc correctly set?

dd slowness is about the deployment progress you mentioned that goes initially really fast and gets sluggish at the 98%-100%. That is normal and is not an indication of anything being wrong.

tdtmusic2 · 2025-01-28T18:53:59Z

Hello @tdtmusic2 Glad to hear finally some progress! Regarding the machines getting into failed deployed even though the actual deployment was fine, I'd say check the cloud-init logs they should have hints as to why the post process failed.

Another Question: Is the machine hostname, IP address etc correctly set?

dd slowness is about the deployment progress you mentioned that goes initially really fast and gets sluggish at the 98%-100%. That is normal and is not an indication of anything being wrong.

You mean the deployed server? Yes, it has the correct hostname and ip, MaaS takes care of those and never had issues.
About the slowness, that's what I was asking about, why do you say is normal? If I download something from somewhere usually it goes full speed all the way, not slowing down at the end :)

iamtinu mentioned this issue Nov 25, 2024

Unable to build Windows Image due to error : Qemu stdout: !!!! X64 Exception Type - 0D(#GP - General Protection) CPU Apic ID - 00000000 !!!! #267

Closed

github-actions bot added the stale label Dec 28, 2024

github-actions bot removed the stale label Jan 10, 2025

SK1Y101 added windows bug Something isn't working labels Jan 23, 2025

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Cannot deploy image after successful build. Is this an issue in windows packer template file. #278

Cannot deploy image after successful build. Is this an issue in windows packer template file. #278

iamtinu commented Nov 25, 2024 •

edited

Loading

MggMuggins commented Nov 25, 2024

nics90 commented Nov 25, 2024

alanbach commented Nov 25, 2024

nics90 commented Nov 26, 2024

iamtinu commented Nov 26, 2024 •

edited

Loading

iamtinu commented Nov 26, 2024 •

edited

Loading

iamtinu commented Nov 26, 2024

iamtinu commented Nov 26, 2024

nics90 commented Nov 26, 2024

iamtinu commented Nov 27, 2024

nics90 commented Nov 27, 2024

iamtinu commented Nov 27, 2024 •

edited

Loading

github-actions bot commented Dec 28, 2024

tdtmusic2 commented Jan 9, 2025 •

edited

Loading

alanbach commented Jan 9, 2025

tdtmusic2 commented Jan 9, 2025

alanbach commented Jan 12, 2025

tdtmusic2 commented Jan 19, 2025

alanbach commented Jan 20, 2025

tdtmusic2 commented Jan 23, 2025

tdtmusic2 commented Jan 23, 2025 •

edited

Loading

alanbach commented Jan 23, 2025

tdtmusic2 commented Jan 23, 2025 •

edited

Loading

tdtmusic2 commented Jan 25, 2025

alanbach commented Jan 27, 2025

tdtmusic2 commented Jan 28, 2025 •

edited

Loading

Cannot deploy image after successful build. Is this an issue in windows packer template file. #278

Cannot deploy image after successful build. Is this an issue in windows packer template file. #278

Comments

iamtinu commented Nov 25, 2024 • edited Loading

MggMuggins commented Nov 25, 2024

nics90 commented Nov 25, 2024

alanbach commented Nov 25, 2024

nics90 commented Nov 26, 2024

iamtinu commented Nov 26, 2024 • edited Loading

iamtinu commented Nov 26, 2024 • edited Loading

iamtinu commented Nov 26, 2024

iamtinu commented Nov 26, 2024

nics90 commented Nov 26, 2024

iamtinu commented Nov 27, 2024

nics90 commented Nov 27, 2024

iamtinu commented Nov 27, 2024 • edited Loading

github-actions bot commented Dec 28, 2024

tdtmusic2 commented Jan 9, 2025 • edited Loading

alanbach commented Jan 9, 2025

tdtmusic2 commented Jan 9, 2025

alanbach commented Jan 12, 2025

tdtmusic2 commented Jan 19, 2025

alanbach commented Jan 20, 2025

tdtmusic2 commented Jan 23, 2025

tdtmusic2 commented Jan 23, 2025 • edited Loading

alanbach commented Jan 23, 2025

tdtmusic2 commented Jan 23, 2025 • edited Loading

tdtmusic2 commented Jan 25, 2025

alanbach commented Jan 27, 2025

tdtmusic2 commented Jan 28, 2025 • edited Loading

iamtinu commented Nov 25, 2024 •

edited

Loading

iamtinu commented Nov 26, 2024 •

edited

Loading

iamtinu commented Nov 26, 2024 •

edited

Loading

iamtinu commented Nov 27, 2024 •

edited

Loading

tdtmusic2 commented Jan 9, 2025 •

edited

Loading

tdtmusic2 commented Jan 23, 2025 •

edited

Loading

tdtmusic2 commented Jan 23, 2025 •

edited

Loading

tdtmusic2 commented Jan 28, 2025 •

edited

Loading