example job link
search to see how prominent the failure is.
It seems to be currently failing in 12% of aws-ovn-upgrade jobs and 8% of openshift-sdn version of that job.

multiple test cases doing similar things are failing around this. If you expand the intervals section of the main job
page you can visually see the disruptions happening through the duration of the job (towards the bottom of the
graph in red)

link to this job's testgrid for reference.