1. Proposed title of this feature request
ROUTER_IP_V4_V6_MODE included on the available possible Ingress operator
override variables.

2. What is the nature and description of the request?
With OCP 3, you could define environment variables on top of the router
deployments, this allowed passing in ROUTER_IP_V4_V6_MODE in case the
underlying systems (routers) were configured to use NodePortService for
exposing haproxy ports.

With OCP 4, that's not the case anymore and ROUTER_IP_V4_V6_MODE is only
ever enabled from the Ingress Operator whenever an IPv6 entry on top of
ClusterNetworkEntry is present as per [1]. For most small/middle scale
deployment enabling IPv6 at the SDN layer is not a general need while
there may be a desire to expose an IPv6 address as taken from the
underlying NodePortService interfaces and similarly to what happens with
IPv4 where ingress routers expose port 80/443 on 80, 443.

There should be a way to override this behavior [2] and allow operators
to define ROUTER_IP_V4_V6_MODE in the eventuality the underlying network
setup was already configured for exposing haproxy ports on IPv6
(::80,::443).

3. Why does the customer need this? (List the business requirements here)
The underlying environment has interfaces configured to also support
IPv6 connectivity. The fact haproxy cannot be configured via
ROUTER_IP_V4_V6_MODE to also bind on IPv6 ports prevents our services
from being reachable on IPv6 as well.

4. List any affected packages or components.
HAproxy

[[1] https://github.com/openshift/cluster-ingress-operator/blob/master/pkg/operator/controller/ingress/deployment.go#L860

[2] https://github.com/openshift/cluster-ingress-operator/blob/master/pkg/operator/controller/ingress/deployment.go#L483