Chapter 6. Completing post customization tasks

To complete the customizations made, perform the following tasks:

  • Create a product.img image file (applies only for graphical customizations).
  • Create a custom boot image.

This section provides information about how to create a product.img image file and to create a custom boot image.

6.1. Creating a product.img file

A product.img image file is an archive containing new installer files that replace the existing ones at runtime.

During a system boot, Anaconda loads the product.img file from the images/ directory on the boot media. It then uses the files that are present in this directory to replace identically named files in the installer’s file system. The files when replaced customizes the installer (for example, for replacing default images with custom ones).

Note: The product.img image must contain a directory structure identical to the installer. For more information about the installer directory structure, see the table below.

Table 6.1. Installer directory structure and custom contents

Type of custom contentFile system location

Pixmaps (logo, sidebar, top bar, and so on.)


GUI stylesheet


Anaconda add-ons


Product configuration files


Custom configuration files


Anaconda DBus service conf files


Anaconda DBus service files


The procedure below explains how to create a product.img file.


  1. Navigate to a working directory such as /tmp, and create a subdirectory named product/:

    $ cd /tmp
  2. Create a subdirectory product/

    $ mkdir product/
  3. Create a directory structure identical to the location of the file you want to replace. For example, if you want to test an add-on that is present in the /usr/share/anaconda/addons directory on the installation system, create the same structure in your working directory:

    $ mkdir -p product/usr/share/anaconda/addons

    To view the installer’s runtime file, boot the installation and switch to virtual console 1 (Ctrl+Alt+F1) and then switch to the second tmux window (Ctrl+b+2). A shell prompt that can be used to browse a file system opens.

  4. Place your customized files (in this example, custom add-on for Anaconda) into the newly created directory:

    $ cp -r ~/path/to/custom/addon/ product/usr/share/anaconda/addons/
  5. Repeat steps 3 and 4 (create a directory structure and place the custom files into it) for every file you want to add to the installer.
  6. Create a .buildstamp file in the root of the directory. The .buildstamp file describes the system version, the product and several other parameters. The following is an example of a .buildstamp file from Red Hat Enterprise Linux 8.4:

    Product=Red Hat Enterprise Linux

    The IsFinal parameter specifies whether the image is for a release (GA) version of the product (True), or a pre-release such as Alpha, Beta, or an internal milestone (False).

  7. Navigate to the product/ directory, and create the product.img archive:

    $ cd product
    $ find . | cpio -c -o | gzip -9cv > ../product.img

    This creates a product.img file one level above the product/ directory.

  8. Move the product.img file to the images/ directory of the extracted ISO image.

The product.img file is now created and the customizations that you want to make are placed in the respective directories.


Instead of adding the product.img file on the boot media, you can place this file into a different location and use the inst.updates= boot option at the boot menu to load it. In that case, the image file can have any name, and it can be placed in any location (USB flash drive, hard disk, HTTP, FTP or NFS server), as long as this location is reachable from the installation system.

6.2. Creating custom boot images

After you customize the boot images and the GUI layout, create a new image that includes the changes you made.

To create custom boot images, follow the procedure below.


  1. Make sure that all of your changes are included in the working directory. For example, if you are testing an add-on, make sure to place the product.img in the images/ directory.
  2. Make sure your current working directory is the top-level directory of the extracted ISO image - e.g. /tmp/ISO/iso.
  3. Create a new ISO image using the genisoimage:

    # genisoimage -U -r -v -T -J -joliet-long -V "RHEL-9 Server.x86_64" -volset "RHEL-9 Server.x86_64" -A "RHEL-9 Server.x86_64" -b isolinux/isolinux.bin -c isolinux/ -no-emul-boot -boot-load-size 4 -boot-info-table -eltorito-alt-boot -e images/efiboot.img -no-emul-boot -o ../NEWISO.iso .

    In the above example:

    • Make sure that the values for -V, -volset, and -A options match the image’s boot loader configuration, if you are using the LABEL= directive for options that require a location to load a file on the same disk. If your boot loader configuration (isolinux/isolinux.cfg for BIOS and EFI/BOOT/grub.cfg for UEFI) uses the inst.stage2=LABEL=disk_label stanza to load the second stage of the installer from the same disk, then the disk labels must match.


      In boot loader configuration files, replace all spaces in disk labels with \x20. For example, if you create an ISO image with a RHEL 9.0 label, boot loader configuration should use RHEL\x209.0.

    • Replace the value of the -o option (-o ../NEWISO.iso) with the file name of your new image. The value in the example creates the NEWISO.iso file in the directory above the current one.

      For more information about this command, see the genisoimage(1) man page.

  4. Implant an MD5 checksum into the image. Note that without an MD5 checksu, the image verification check might fail (the option in the boot loader configuration) and the installation can hang.

    # implantisomd5 ../NEWISO.iso

    In the above example, replace ../NEWISO.iso with the file name and the location of the ISO image that you have created in the previous step.

    You can now write the new ISO image to physical media or a network server to boot it on physical hardware, or you can use it to start installing a virtual machine.

Additional resources