When using an installer with amd64 payload, configuring the VM to use aarch64 is possible through the installer-config.yaml:

additionalTrustBundlePolicy: Proxyonly
apiVersion: v1
baseDomain: ci.devcluster.openshift.com
compute:
- architecture: arm64
  hyperthreading: Enabled
  name: worker
  platform: {}
  replicas: 3
controlPlane:
  architecture: arm64
  hyperthreading: Enabled
  name: master
  platform: {}
  replicas: 3

However, the installation will fail with ambiguous error messages:

ERROR Attempted to gather ClusterOperator status after installation failure: listing ClusterOperator objects: Get "https://api.build11.ci.devcluster.openshift.com:6443/apis/config.openshift.io/v1/clusteroperators": dial tcp 13.59.207.137:6443: connect: connection refused

The actual error hides in the bootstrap VM's System Log:

Red Hat Enterprise Linux CoreOS 417.94.202407010929-0 4.17

SSH host key: SHA256:Ng1GpBIlNHcCik8VJZ3pm9k+bMoq+WdjEcMebmWzI4Y (ECDSA)

SSH host key: SHA256:Mo5RgzEmZc+b3rL0IPAJKUmO9mTmiwjBuoslgNcAa2U (ED25519)

SSH host key: SHA256:ckQ3mPUmJGMMIgK/TplMv12zobr7NKrTpmj+6DKh63k (RSA)

ens5: 10.29.3.15 fe80::1947:eff6:7e1b:baac

Ignition: ran on 2024/08/14 12:34:24 UTC (this boot)

Ignition: user-provided config was applied

[0;33mIgnition: warning at $.kernelArguments: Unused key kernelArguments[0m



[1;31mRelease image arch amd64 does not match host arch arm64[0m

ip-10-29-3-15 login: [   89.141099] Warning: Unmaintained driver is detected: nft_compat

    

4.16
    

Use amd64 installer to install a cluster with aarch64 nodes
    

    1. download amd64 installer
    2. generate the install-config.yaml
    3. edit install-config.yaml to use aarch64 nodes
    4. invoke the installer
    

installation timed out after ~30mins
    

installation failed immediately with proper error message indicating the installation is not possible
    

https://redhat-internal.slack.com/archives/C68TNFWA2/p1723640243828379