A vK8S namespace has different DNS suffix depending on if the POD is hosted on a RE or a CE. The expected behavior is the DNS suffix engineering in the RE: to have the tenant and namespace info, but do not have the CE site name.

Could we have the same DNS suffix in RE and CE?

Impact: A NGINX cluster of instances is synchronizing their zones between NGINX instances. A K8S headless service is created and used by NGINX to synchronize instances. NGINX is not using the "search", so the FQDN is set in NGINX config. For example: my-headless-service-name.my-tenant--my-namespace.svc.cluster.local . Due to CE behavior (different DNS suffix per CE site), the NGINX hosted in CEs can't share the same config, that is not compliant with NGINX One.

Current behavior, the line "search" in /etc/resolv.conf:
on RE: my-tenant--my-namespace.svc.cluster.local

on CE: my-namespace.svc.my-site.my-tenant.tenant.local