With the 200Mi memory limit for the whereabouts daemon-set the POD is getting OOMKilled in large clusters (100+ workers).
We locally tested with limit set to 400Mi and the issue is not seen.
We run the daemon sets as so,
command: [ "/bin/sh" ]
args:
- -c
- >
SLEEP=false /install-cni.sh &&
/ip-control-loop -log-level debug

So the PODs are not sleeping after the install. ip-control-loop is running continuously.

Raising this ticket to consider that change.