nginx-ingress controller upgrades - aws load balancer detached issues

[suranga-jayalath]

I upgraded the Ingress Controller from version v4.12.0 to v4.12.4 using the Helm chart with a rolling update strategy. All the Ingress Controller pods came up successfully without any errors.

However, after the upgrade completed, I noticed that other Ingress services (attached to the AWS Load Balancer) were detached and then reattached after around 40 seconds.

What could be the reason for this behavior?
Is there any way to prevent this and ensure zero downtime for traffic during the upgrade?

[vepatel]

hey @suranga-jayalath, i think you might be using a different project https://github.com/kubernetes/ingress-nginx/releases which is k8s community ingress controller not this one