User Story:

When a non-cluster (internet) DNS query is forwarded by coreDNS to an upstream nameserver, the destination should be configured to be AWS Route53 Resolver Service at fd00:ec2::253 (AWS IPv6 name server).  If using custom DNS, the default nameserver should also be configured to be fd00:ec2::253.

The reason is that pods will only have IPv6 and the node's /etc/resolv.conf only defines an IPv4 namesever. Thus, the query will fail due to "unreachable network" error.

Acceptance Criteria:

Description of criteria:

• DNS operator specs define fd00:ec2::253 as an upstream namesever for the coreDNS. 
• When custom DNS is enabled, unit file that updates cluster DNS should be configured to set fd00:ec2::253.

(optional) Out of Scope:

Detail about what is specifically not being delivered in the story

Engineering Details:

• Below is an example manifest that configure the DNS operator's upstream nameservers.

apiVersion: operator.openshift.io/v1
kind: DNS
metadata:
  name: default
spec:
  upstreamResolvers: 
    policy: Sequential
    upstreams:
    - type: Network
      address: fd00:ec2::253
    - type: SystemResolvConf 

• Unit file that updates cluster DNS (when custom DNS is enabled): https://github.com/openshift/machine-config-operator/blob/main/templates/common/aws/units/aws-update-dns.service.yaml 

This requires/does not require a design proposal.
This requires/does not require a feature gate.