Right now e2e tests for primary-UDN and secondary interfaces are basically the same, with a few diffs:
1. networkInterfaceName is different (nad-name vs "ovn-udn1)
2. VMI is created with different interface (secondary vs primary binding )
3. NAD created with different role(secondary vs primary)
4. IPs retreival: In Secondary the IPs are retrieved via vmi.status, while in primaryUDN the IPS are extracted using network-status annotation.

1-3 can be parameterized to achieve cleaner code.
4 should be aligned to vmi.status as soon as kubevirt/kubevirt supports it for primaryUDN.