The recent addition in 4.8 of single-node cluster support via https://github.com/openshift/cluster-etcd-operator/pull/527 contains an absolute image reference to the upstream origin-cli image. Not only will this end up pulling an upstream image even in a downstream OCP build, it causes etcd-quorum-guard to crash on IBM P&Z architectures because upstream is amd64-only. This needs to be fixed to respect the substitution of 'cli' from the payload.
specifically this commit: https://github.com/openshift/cluster-etcd-operator/pull/527/commits/0433d66c6011b58936db4712fa829e6686ef30fd
nice catch yes when we moved the spec out of CVO control we did not remember to populate the image pull-spec from the release payload.
verified w.r.t multi-arch on s390x with https://mirror.openshift.com/pub/openshift-v4/s390x/clients/ocp-dev-preview/4.8.0-0.nightly-s390x-2021-02-17-131252/ and the cli image used is the downstream image for OCP.
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory (Moderate: OpenShift Container Platform 4.8.2 bug fix and security update), and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHSA-2021:2438