Avg-normal_latency and 99 percentile for udp protocol and message size=1024kb  are degraded by more than 100% in 4.12 network perf tests when compared to the 4.11 network perf tests. 
Cloud provider is NUTANIX
The test type is RR [request and response]

Version-Release number of selected component (if applicable): 4.12.0-0.nightly-2022-12-17-123606
 (openshift cluster aws)How reproducible:
Seen in 2 back to back tests and the issue was also reported earlier for sdn
reference bugzilla-->https://bugzilla.redhat.com/show_bug.cgi?id=2076964 
https://issues.redhat.com/browse/OCPQE-13230

Steps to Reproduce:
1.install  4.12 and 4.11.z ocp clusters through Flexy
2.Run a network perf pod2pod test[uperf] for the 4.11.z and 4.12 clusters
3.Compare the network perf results 

Actual results:
The pod2pod network perf results are degraded for "99-percentile-average-latency" and "avg normal-latency" by more than 100% for 4.12 tests when compared to 4.12 network perf tests.


Expected results:
The pod2pod network perf results should not be deviated/degraded by more than 20-25% when compared to the baseline[4.11.z in this case]

Additional info:
Networking workload targeted  pod2pod.pod2svc and hostnet

Google spreadsheet link with the bar charts and all the reports
- https://docs.google.com/spreadsheets/d/1mglDFYAwirzJhew3Q8A7bRybGjHtPHCk12t8WD7UrhQ/edit#gid=0


The tests were launched at the same time to reduce the impact of unstable provider network conditions and/or noisy neighbors.Network perf tests are run from here; https://mastern-jenkins-csb-openshift-qe.apps.ocp-c1.prod.psi.redhat.com/job/scale-ci/job/e2e-benchmarking-multibranch-pipeline/job/network-perf/https://github.com/cloud-bulldozer/e2e-benchmarking/tree/master/workloads/network-perfThe purpose of the network scripts is to run uperf workload on the Openshift Cluster


benchmark-operator will be the namespace created by the load-test; client and server pods are pods created under this ns.workload  is uperf 

https://github.com/cloud-bulldozer/benchmark-operator/blob/master/docs/uperf.md

to run the network perf test manually 
https://docs.google.com/document/d/1oLY_OluT_VqYD8x1oK-hF6ScBwL8sKG8DLn53PpEX8o/edit