AWS edge zone jobs are failing in cloud-provider-aws (CCM-AWS) tests when the selected (compute/worker) node is not allowed to schedule.

cloud-provider-aws tests are exposed through OTE, edge-zone jobs are unscheduled by design, requiring workloads to tolerate to intentionally run workloads on top of it, that's not the case of those tests

Considering the nodes is listed alphabetically, those edge zones is also labeled with node-role.kubernetes.io/worker, it is selected without checking the existence of node taints, as well without setting required tolerations:
  taints:
    - effect: NoSchedule
      key: node-role.kubernetes.io/edge

Tests that we are observing impacting those jobs:
[cloud-provider-aws-e2e] loadbalancer NLB internal should be reachable with hairpinning traffic [Suite:openshift/conformance/parallel]
[cloud-provider-aws-e2e] loadbalancer CLB internal should be reachable with hairpinning traffic [Suite:openshift/conformance/parallel] 

4.22+    

    always when the node is first in the list of worker nodes (which is listed by hostname)

    1.
    2.
    3.
    

1. cloud-provider-aws tests selects primarily worker nodes that can be scheduled

Example: jobs failing:
    https://prow.ci.openshift.org/view/gs/test-platform-results/logs/periodic-ci-openshift-release-main-nightly-4.22-e2e-aws-ovn-edge-zones/2025360636724121600