Description of problem: The Machine Suite: [It] Autoscaler should use a ClusterAutoscaler that has 8 maximum total nodes scales up and down while respecting MaxNodesTotal [Slow][Serial] [autoscaler, periodic] test fails for OpenStack periodics in the 4.13, 4.14, and 4.15 branches where we're using compact clusters. The test hardcodes the number of machines it expects at the beginning, which is 3 and not 6 for compact clusters.

For some obscure reason, the 4.16 and master periodics are setup to use full blown clusters and are thus not affected. We however want to move these to compact cluster to save resources.

The test fails with:

{Timed out after 1800.001s.
Expected to have 6 machines in the cluster
Expected
    <int>: 3
to be equivalent to
    <int>: 6 failed [FAILED] Timed out after 1800.001s.
Expected to have 6 machines in the cluster
Expected
    <int>: 3
to be equivalent to
    <int>: 6
In [It] at: /go/src/github.com/openshift/cluster-api-actuator-pkg/pkg/autoscaler/autoscaler.go:609 @ 06/16/24 20:19:42.333
}

Example failure: https://prow.ci.openshift.org/view/gs/test-platform-results/logs/periodic-ci-openshift-cluster-api-actuator-pkg-release-4.13-e2e-openstack-operator-periodic/1802407569927442432

Version-Release number of selected component (if applicable):

How reproducible:

Steps to Reproduce:

1.
2.
3.

Actual results:

Expected results:

Additional info: