RamLavi

/cherry-pick release-0.45

/test pull-kubemacpool-e2e-k8s

I also included bumps to the latest versions suggested by renovate.

/test pull-kubemacpool-e2e-k8s just me thriple checking the e2e is OK

[Change](https://github.com/ovn-kubernetes/ovn-kubernetes/compare/8d8b9bd04c89960e6017d48f110a6dd47748b88b..943a8ecf0f38814120ddb2e60988bfd73713f3ee): Address @qinqon's review

[5/5 passes](https://github.com/ovn-kubernetes/ovn-kubernetes/actions/runs/19294560339/job/55176368542?pr=5719) on dual stack live-migration lane [Skipped](https://github.com/ovn-kubernetes/ovn-kubernetes/actions/runs/19294560339/job/55176368506?pr=5719) on ipv4 live-migration lane (don't we want to enable static ips `ENABLE_PRE_CONF_UDN_ADDR` FG there?

[Change](https://github.com/ovn-kubernetes/ovn-kubernetes/compare/943a8ecf0f38814120ddb2e60988bfd73713f3ee..c666a48de4eba8d96211beb71dab963e2130290f): Rebase (no conflicts)

> e2e failures look legit ah, yes. the test fails are indeed legit, I will need to refactor the existing duplicate IPs tests to make them check it more accurately.

[Change](https://github.com/ovn-kubernetes/ovn-kubernetes/compare/c666a48de4eba8d96211beb71dab963e2130290f..76eb7c5c6ed3c642fc28884429f11ab064b00922): Addressed @kyrtapz review, refactored e2e tests that failed due to the current change

Output of POC looks like: ``` [root@zeus33 ovn-kubernetes]# kubectl get pod -n udn-primary-gateway NAME READY STATUS RESTARTS AGE virt-launcher-worker2-zr9j9 2/2 Running 0 17m ``` `k8s.ovn.org/pod-networks` annotation: ``` [root@zeus33 ovn-kubernetes]# kubectl...