OOM killed because of rollout restart

[karantyagi-cashfree]

Hi all,
Every time we perform a rollout restart of our Ingress controller (currently on version 1.11.5), we observe 4–5 times of OOMKilled events before it stabilises. After these initial failures, the controller eventually runs without issues.

[haywoodsh]

Hi @karantyagi-cashfree, I'm afraid we didn't have the release 1.11.5. Can you confirm that you are indeed deploying NGINX Ingress Controller. If you are using the community ingress controller, please refer to https://github.com/kubernetes/ingress-nginx for more information.