Description of problem:

An existing installation of RHODS had a complete break of the `odh-model-controller` pods after upgrading to OCP 4.13, and it looks like the root cause is the clusterrole being used (`manager-role`) which is an example one and used by RHACM as well, which might have caused the issue to RHODS during its own upgrade.
RHODS should use a different clusterrole to avoid any further issues.

More details: https://redhat-internal.slack.com/archives/C03UGJY6Z1A/p1684747070780669

Prerequisites (if any, like setup, operators/versions):

RHODS + RHACM (or any other project that uses the same clusterrole)

Steps to Reproduce

Unclear, but likely an upgrade to RHACM overrode the clusterrole that the odh-model-controller pods are using

Actual results:

odh-model-controller pods keep restarting

Expected results:

No issues caused by other projects

Reproducibility (Always/Intermittent/Only Once):

Happened only once that we know of, likely can happen again.

Build Details:

Workaround:

Additional info: