Jump to contentJump to page navigation: previous page [access key p]/next page [access key n]
SUSE Edge Documentation / Tips and Tricks / Edge Image Builder

26 Edge Image Builder

Common
  • If you are in a non-Linux environment and following these instructions to build an image, then you are likely running Podman via a virtual machine. By default, this virtual machine will be configured to have a small amount of system resources allocated to it and can cause instability for Edge Image Builder during resource intensive operations, such as the RPM resolution process. You will need to adjust the resources of the podman machine, either by using Podman Desktop (settings cogwheel → podman machine edit icon) or directly via the podman-machine-set command

  • At this point in time, the Edge Image Builder is not able to build images in a cross architecture setup, i.e. you have to run it on:

    • aarch64 systems (such as Apple Silicon) to build SL Micro aarch64 images

    • x86_64 systems (such as Intel) to build SL Micro x86_64 images.

Kubernetes
  • Creating multi node Kubernetes clusters requires adjusting the kubernetes section in the definition file to:

    • list all server and agent nodes under kubernetes.nodes

    • set a virtual IP address that would be used for all non-initializer nodes to join the cluster under kubernetes.network.apiVIP

    • optionally, set an API host to specify a domain address for accessing the cluster under kubernetes.network.apiHost To learn more about this configuration, please refer to the Kubernetes section docs.

  • Edge Image Builder relies on the hostnames of the different nodes to determine their Kubernetes type (server or agent). While this configuration is managed in the definition file, for the general networking setup of the machines we can utilize either DHCP or the Edge Networking page.