Hi all, Our EU OS image build based on ublue version 44 fails with the error:

[08:29:52 r.g.c/e/w/e/eu-os-demo:ublue-44] => #49 exporting to image
[08:29:52 r.g.c/e/w/e/eu-os-demo:ublue-44] => #49 pushing layers 50.2s done
[08:29:52 r.g.c/e/w/e/eu-os-demo:ublue-44] => #49 pushing manifest for registry.gitlab.com/eu-os/workspace-images/eu-os-base-demo/eu-os-demo:ublue-44@sha256:c28adf5789ed495ecc7c2d31ada7ebf896d6c2c09f762f0b241e75ab325e8987
[08:29:52 r.g.c/e/w/e/eu-os-demo:ublue-44] => #49 pushing manifest for registry.gitlab.com/eu-os/workspace-images/eu-os-base-demo/eu-os-demo:ublue-44@sha256:c28adf5789ed495ecc7c2d31ada7ebf896d6c2c09f762f0b241e75ab325e8987 0.0s done
[08:29:52 r.g.c/e/w/e/eu-os-demo:ublue-44] => #49 ERROR: failed to push registry.gitlab.com/eu-os/workspace-images/eu-os-base-demo/eu-os-demo:ublue-44: unknown: too many manifest references
[08:29:52 r.g.c/e/w/e/eu-os-demo:ublue-44] => 
[08:29:52 r.g.c/e/w/e/eu-os-demo:ublue-44] => #51 exporting cache to registry
[08:29:52 r.g.c/e/w/e/eu-os-demo:ublue-44] => #51 sending cache export 46.0s done

Source: https://gitlab.com/eu-os/workspace-images/eu-os-base-demo/-/jobs/16610306184 Any idea what's wrong? Our recipe file is here: https://gitlab.com/eu-os/workspace-images/eu-os-base-demo/-/blob/main/recipes/ublue-44.yml?ref_type=heads

message has been deleted

I'm using the ghcr.io/blue-build/base-images/fedora-silverblue image which seems to come with opencascade-draw included for some reason. dnf list shows this Installed packages (available for reinstall, available for upgrade) opencascade-draw.x86_64 7.9.3-4.fc44.1 updates opencascade-foundation.x86_64 7.9.3-4.fc44.1 updates opencascade-modeling.x86_64 7.9.3-4.fc44.1 updates opencascade-ocaf.x86_64 7.9.3-4.fc44.1 updates opencascade-visualization.x86_64 7.9.3-4.fc44.1 updates so I thought this would fix it remove: packages: - opencascade-draw in the action I get No packages to remove for argument: opencascade-draw any ideas?

New to this, how do I add the proprietary Broadcom wifi driver (wl)?

G'day, I've only just gotten started using BlueBuild so this seems like a silly question, and I can't seem to find any proper (at least, up to date) tutorial/documentation that goes over this step specifically, but For the longest time now, for my repo I've had to upload both my MOK.der and MOK.priv keys publicly to get successful builds. Naturally, I shouldn't do this, and I should make the MOK.priv key a repo secret so it builds without needing to upload the private key file. However, I'm unsure what to name the secret, or how to edit any scripts so that the build action reads the secret. Making secrets named MOK_PRIV and SECURE_BOOT_PRIV_KEY, with the MOK.priv key inside, doesn't do anything. Again, this may be completely trivial and not appropriate for this channel, and I apologise if that's the case, I'm just looking for help with the one step I haven't completely wrapped my head around yet.

Hello, I was trying to make an image that includes steam, but I cannot rebase to it, this is the error I have been getting when running `rpm-ostree rebase ostree-image-signed:docker://ghcr.io/ceruleanderpo/kinoite-nvidia-gaming:latest`:

error: Updating rpm-md repo 'fedora-rar-debug': cannot update repo 'fedora-rar-debug': Cannot download repomd.xml: Cannot download repodata/repomd.xml: All mirrors were tried; Last error: Status code: 404 for https://negativo17.org/repos/rar/fedora-44/x86_64.debug/repodata/repomd.xml (IP: 51.15.18.69)

These are the recipe files: 1.

recipe-gaming.yml

---
# yaml-language-server: $schema=https://schema.blue-build.org/recipe-v1.json
# image will be published to ghcr.io/<user>/<name>
name: kinoite-nvidia-gaming
# description will be included in the image's metadata
description: A version of kinoite-nvidia that includes Steam

# the base image to build on top of (FROM) and the version tag to use
base-image: ghcr.io/ublue-os/kinoite-main
image-version: latest # latest is also supported if you want new updates ASAP

# module configuration, executed in order
# you can include multiple instances of the same module
modules:
  - from-file: common-modules.yml

  - type: dnf
    repos:
      nonfree: negativo17
    install:
      skip-unavailable: true
      packages:
        - steam

I got the following error when doing a

rpm-ostree upgrade

error: ostree-tar: Failed to handle file: ostree-tar: Failed to import file: Writing content object: min-free-space-percent '3%' would be exceeded

I had to run a

sudo rpm-ostree cleanup -r

to get enough space for an upgrade. I suspect

boot

being 2GB is part of the problem? This seems to be the default config in the installer:

nvme0n1                                       259:0    0 238.5G  0 disk
├─nvme0n1p1                                   259:1    0   600M  0 part  /boot/efi
├─nvme0n1p2                                   259:2    0     2G  0 part  /boot
└─nvme0n1p3                                   259:3    0 235.9G  0 part
  └─luks-edde129a-a421-4589-b8b5-44fed5d62c12 252:0    0 235.9G  0 crypt
    └─base-root                               252:1    0    15G  0 lvm   /var
                                                                         /sysroot/ostree/deploy/default/var
                                                                         /sysroot
                                                                         /etc

I can reinstall on this machine to change the sizes (currently a test deployment), but was wondering how big the boot partition should be here. I do not know if luks is causing the sizes to be off here.

At the moment I need to disable secure boot, enroll the key with mokutil and then reenable secure boot. I was wondering if there was a way to make the installer call mokutil with the key in the image. Of course this does not prevent a user from forgetting to do this and needing to disable secure boot.

Hey y'all, Im pretty new to this but I installed FCOS using [their docs](https://docs.fedoraproject.org/en-US/fedora-coreos/bare-metal/), then had a pretty basic butane file (below), rebooted and FCOS looked fine. I then ran

bootc switch ghcr.io/blue-build/base-images/fedora-cosmic-nvidia

and when i rebooted saw that the

ostree-boot-complete

service failed. The log said it was out of space somehow 🤔 I ran a

df -h

and everything looks fine butane:

variant: fcos
version: 1.4.0
passwd:
  users:
    - name: matt
      ssh_authorized_keys:
        - <redacted>
      password_hash: "<redacted>"
      groups:
        - wheel
        - sudo
        - docker

storage:
  files:
    - path: /etc/hostname
      mode: 0644
      contents:
        inline: ludd
systemd:
  units:
    - name: docker.socket
      enabled: true
      contents: |
        [Unit]
        Description=Docker Socket for the API
        PartOf=docker.service

        [Socket]
        ListenStream=/var/run/docker.sock
        SocketMode=0660
        SocketUser=root
        SocketGroup=docker

        [Install]
        WantedBy=sockets.target

hey! im very new to this, but im wonderin is there any maintenance required for a bluebuild repo other than updating the base image and adjusting any changes i made as need?

normally, you'd set the

accent-color

gsetting on the

gdm

user on a normal distro. how do i do this in bluebuild the nice way?

the tutorial on how to build an iso makes non netinst isos, how would i make a netinst iso (and, if possible, make it happen automatically with github actions?) The "How to set up a new repository" docs page sys that it will teach you how to make a netinstall ISO, but it never does. https://cdn.discordapp.com/attachments/1526728598151499980/1526728598302752980/image.png?ex=6a58144f&is=6a56c2cf&hm=ff8c318ae07c99b6fdcd3513d18d4abc53f4269e8bda3739be131cc688f43a6d&

I have been trying to figure out user creation when basing off a

quay.io/fedora-ostree-desktops/base-atomic

image. Running

bluebuild generate-iso --iso-name cerberus-desktop.iso recipe recipes/recipe-cerberus.yml

will generate an installer but this installer will not ask for user creation. From what I can tell, this is because the wrong image variant for generated iso but I see no way to force the specific variant? I feel I am missing something completely here regarding initial user creation. The image is otherwise entirely empty other than sddm since I haven't done other bits of install yet. I tried following what wayblue does for sddm setup but I suspect this is not going to solve my issues after trying it. I could see needing to include a desktop to make anything happen here. repository: https://github.com/i509VCB/verdigris

Hi all, ISOs generated with the standard generate iso command from bluebuild are unable to restart when installation is complete. The installer shows a restart button at the end that only leads to a white screen. I tested it across different installs but its always the same. Any other experiemental or suggested ways to build an ISO for release? I'm willing to put the effort into testing and reporting back.

I'm getting this error message when trying to build:

ERROR: failed to build: failed to solve: ghcr.io/ublue-os/akmods-extra:bazzite-44: failed to resolve source metadata for ghcr.io/ublue-os/akmods-extra:bazzite-44: ghcr.io/ublue-os/akmods-extra:bazzite-44: not found

I'm trying to use the Bazzite kernel and add some akmods to it but it says that it cannot find something

Using 1password via the bling module results in an installed 1password (and op), and they can be started from the terminal (starting 1password adds to the tray as well in KDE), but if you open it via the launcher there's a popup "Could not find the program: /opt/1Password/1password"

ls -alh /opt/1Password

shows the symlink to

/usr/lib/opt/1Password

but the link is broken since there is no

/usr/lib/opt/1Password

This is built on

quay.io/fedora-ostree-desktops/kinoite:44

. I had previously installed 1Password via the

dnf

module which seemed to work (although I didn't attempt to login or anything), but I swapped when I saw the bling script considering it does quite a few extra things and I wasn't sure if those were necessary. My original goal here was to set up 1Password natively and Firefox via flatpak with a default override allowing Firefox to use the 1Password IPC for browser integration (which, for all I know, would work out of the box if I'd actually just try it).

When using the Github action of blue-build, how can I postprocess the container image, e.g. do additional build steps? I can select

push: false

, but none of

podman/docker/buildah images

lists the image created by the blue-build step in the local image store. Where is the image stored within the runner, so I can further work with it locally before pushing to the registry?

I'm running into weird behavior where some packages are not available for install during the dnf module of blue-build, but are part of fedora's repos and are installable via rpm-ostree. An example of such a package is

polkit-qt6-1-devel

(a dependency of a copr package I maintain). When I try and install it directly, I get

l/crafti-hyprland-rpm-prerelease-track:latest_linux_amd64 => No match for argument: polkit-qt6-1-devel

and when I try to add a package that depends on it, I get

[16:14:36 g.i/c/crafti-hyprland-rpm-prerelease-track:latest] => 11.87  Problem 2: conflicting requests
[16:14:36 g.i/c/crafti-hyprland-rpm-prerelease-track:latest] => 11.87   - nothing provides libpolkit-qt6-agent-1.so.1()(64bit) needed by hyprpolkitagent-0.1.3-5.fc44.x86_64 from copr:copr.fedorainfracloud.org:craftidore:hyprland-prerelease-track
[16:14:36 g.i/c/crafti-hyprland-rpm-prerelease-track:latest] => 11.87   - nothing provides libpolkit-qt6-core-1.so.1()(64bit) needed by hyprpolkitagent-0.1.3-5.fc44.x86_64 from copr:copr.fedorainfracloud.org:craftidore:hyprland-prerelease-track

Adding

allow-erasing

does not solve the error, it seems the package genuinely doesn't exist as far as bluebuild is concerned. But as stated before, I can install it with rpm-ostree just fine. - [x] I have searched this forum for answers to this question - [x] I have read the docs on the dnf module, in case it was answered there

In my image i'm deleting Konsole from KDE, but on launch it still shows a (non-existent file) on the taskbar. I wanted to delete that. How do I do this with a script?

I think I have a pretty quick machine for doing a local build (FW desktop), but I still spend most of my time waiting for package downloads and installs in a build. KDE and ZFS tend to be the two things that take the longest. I was wondering if it was possible to allow caching the changes of these two things so that any updates on top are quicker.

message has been deleted

Hello, I'm enjoying the bootc, blue-build, and even the ublue, items on bootc. I created my own (for learning experiences) image and I really want to dig more and learn more. One thing I want to further do is focus on how to custom format my drive and use different file systems, one of many things. With all that said... tldr: What are some good resources to learn more about Bootable Containers (bootc), containers, and cloud native infra in general? Thanks!

like based on fedora arm64? sorry if its a silly question

so for the past month or so, i haven't been able to

bluebuild switch

to my image locally. i'm only (and have only) used local images, and it worked previously. building works, but when i try switching (

sudo bluebuild switch ./recipes/recipe.yml

), i get this after all the blob copy sha256 lines:

l/ivy-ublue-gnome:latest => Writing manifest to image destination
l/ivy-ublue-gnome:latest => Loaded image: localhost/ivy-ublue-gnome:latest
l/ivy-ublue-gnome:latest => note: automatic updates (stage) are enabled
l/ivy-ublue-gnome:latest => Pulling manifest: ostree-unverified-image:containers-storage:localhost/ivy-ublue-gnome:latest
l/ivy-ublue-gnome:latest => error: Preparing import: Fetching manifest: failed to invoke method OpenImage: reference "[overlay@/var/lib/containers/storage+/run/containers/storage:overlay.imagestore=/usr/lib/containers/storage,overlay.mountopt=nodev,metacopy=on]localhost/ivy-ublue-gnome:latest" does not resolve to an image ID
ERROR => Failed:
  × Failed to switch to image localhost/ivy-ublue-gnome:latest

i've tried disabling signing but to no avail, i'm still pretty new to ublue and bluebuild and all this so yeah :3

Hello Im trying to install protonvpn using official repo however this error occurs i know that it wants booted system is there any workaround to use protonvpn (excluding flatpak)

Is there a way to pin the signing keys for the base image and ensure any packages installed by the module are verified? I'd like to verify all the packages that I can.

Is there a way to pin the signing keys for the base image and ensure any packages installed by the module are verified? I'd like to verify all the packages that I can.

I’m trying to push to a repo hosted with a self signed certificate. How would I trust the certificate in the bluebuild cli so that the container can access that repo?

https://github.com/JasonN3/build-container-installer/wiki/examples_adding-flatpaks jasonN3 installer provides an option for it. Is it possible with the bluebuild cli?

message has been deleted